{"id": "train_045000", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n cost = 0\n for _ in range(budget):\n p = plan(issue)\n patch = edit(issue, p)\n ok, _ = test(patch)\n cost += 1\n if ok:\n return patch, cost\n raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["agent", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "bbb2b129d49994dd77d74f8eec7dc9e629bbff42"} {"id": "train_045001", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "3fdc778ba74a40956460121f691da91f50e39416"} {"id": "train_045002", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["research", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "9c49760a46df9b0712e5920723d25b283042ca11"} {"id": "train_045003", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "9ded81be5092ef8d0853b5cc10cc9a8aa6f9275c"} {"id": "train_045004", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["governance", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "eefd915c00b15604ab030238ff11a87bd12c6731"} {"id": "train_045005", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["coding", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "47806d60e3dfff0505df5dcacea1b8835b613da8"} {"id": "train_045006", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "2c87e357350b91fe09c0bb63379e0204f9712063"} {"id": "train_045007", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["agent", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "dda705ba86483b928221c899e171022ede85b79e"} {"id": "train_045008", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Rust", "moe_experts": ["security", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "7c90998d0030a3e0e4ba65f6c0706a342a8b45e6"} {"id": "train_045009", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "ea797ea0e5deb124467dcab099129af29fd1a83b"} {"id": "train_045010", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "9c567b5e7942d76fd0960765a53f9a2b2f749051"} {"id": "train_045011", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "76b26a2938ee9f394c7a89500838ac27a453c766"} {"id": "train_045012", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["agent", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "ffcb8faf9beff9d1422eaabf7ac9f9daf994ec13"} {"id": "train_045013", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["security", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "fc846ac51d3912c02579fb1d3ad443a9b7dcacc4"} {"id": "train_045014", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["coding", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "fe2bd2e1ba29d6e619955de07be50574c5f5962f"} {"id": "train_045015", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["research", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "b6648a3c2387545e8d5e479ac86df60acdf0fe36"} {"id": "train_045016", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 9,\n \"confidence\": 0.9,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["performance", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "bd22f48b5b8d8bb19b097ed3ad285c932981e570"} {"id": "train_045017", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "SQL", "moe_experts": ["coding", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "17ec10035d61c8e3730e8730d8b5c81c97bac588"} {"id": "train_045018", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "a4cc70a84051cdb4631282494dbaa564ac28c1fa"} {"id": "train_045019", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Python", "moe_experts": ["evaluation", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "96fb2126997aaaf861a0db370e83e798d1339f0c"} {"id": "train_045020", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["security", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "f39f0a2b80a51a3e4d63145da2ba738b191a66b1"} {"id": "train_045021", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "95fc64ba2f567bab0039d3a128bacd26726c8802"} {"id": "train_045022", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["coding", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "fc7975e44efbf0f26a5d72463b5503c36a51a563"} {"id": "train_045023", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "d369183bd921d964406216d732765eb96d1ad85a"} {"id": "train_045024", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["coding", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "96e850d35dd523a9a9854536406d3c260db14fc3"} {"id": "train_045025", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "26e50be306689d348d6dd4994d5b184d7a69ebbc"} {"id": "train_045026", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "ac59ff758f563480600e2d00879c89367ed5e5d6"} {"id": "train_045027", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "13edaf79ad3797b21ff9ec603dd00ddba53eb05b"} {"id": "train_045028", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["coding", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "71aed899ef96322ccb076ed561ca27b8bb0c39f9"} {"id": "train_045029", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "b02888d2fb2af5bd5eac9502908d110cb2111efa"} {"id": "train_045030", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "0f79bc154f88f6e91a2d7a6dc66b9b209ca2e895"} {"id": "train_045031", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["evaluation", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "5a48caa0e37b38adc3b8a4ba4012f6b5474bbd41"} {"id": "train_045032", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["coding", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "ca1f87785d73e5d398ff39fbc33f26823649e85d"} {"id": "train_045033", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["data", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "8b85157d14bc5884a82789738187020b7b90f906"} {"id": "train_045034", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["security", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "ecfa247004eef55c45e18636681eac75cfbde7f6"} {"id": "train_045035", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "c0d29c5da637eeafe0ab53f6c83c0b5eff8ef570"} {"id": "train_045036", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "23a004a4bae03966bcec673806959b305d5dd5d6"} {"id": "train_045037", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["agent", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "fa78ff8da5f99d037ed78cfcfdb10b01ea6eb783"} {"id": "train_045038", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Go", "moe_experts": ["agent", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "0148fca5e5862d7b243a161d211d1b0c4f2969ee"} {"id": "train_045039", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["performance", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "62ff9e637564241a849cb7e66a19e2ce08b8b02c"} {"id": "train_045040", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["agent", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "9b37cebd4811c9674cee0f55b75672a18c9322e1"} {"id": "train_045041", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "cd97c5dddd067bf7f014de5de57d32a3d2929770"} {"id": "train_045042", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["research", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "8bc41e84b3de6b137c1e41454bfd43cf91b1e1af"} {"id": "train_045043", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["security", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "bf5d4d796973dc1cdc39373f1ec4a69d0fdc896e"} {"id": "train_045044", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["agent", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "cdaddb9931dc6581a22fbf1267be3099d1224630"} {"id": "train_045045", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["data", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "af45330e06236d5d9be8dfe5a34f0c0c699e0acc"} {"id": "train_045046", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["coding", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "87303f74e50a10640cc6efe741949b831fc22522"} {"id": "train_045047", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["performance", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "bbbc1241dacdeffc06b77fc010200163d9fe16a0"} {"id": "train_045048", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["data", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "58b505483d587aa23e4fcb6635d957215c6bd9e3"} {"id": "train_045049", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["performance", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "bd02296d19af3d2c09b447bdd79f67ceea15001f"} {"id": "train_045050", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["security", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "beabb263c9bf6ada4a73be63f786d82557f96ca2"} {"id": "train_045051", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "b3822b739dd24310ccd157d2961dfc6a51a40932"} {"id": "train_045052", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "e99f67ca7819208100ff1cb575c744f0b0b3be89"} {"id": "train_045053", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Python", "moe_experts": ["agent", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "98d5f0d96eca6acfc75309cacad790e083da8ed9"} {"id": "train_045054", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "7b4dcb092db02f18108c9448f028964f16c20718"} {"id": "train_045055", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["research", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "729e36e1df6b4901d9cfca662807541eccf5bb96"} {"id": "train_045056", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["coding", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "63426b88d3f23c647505838446c6d14fed32005c"} {"id": "train_045057", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["research", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "c4744a75e59eeef1081ab79682de0ccd86c44b19"} {"id": "train_045058", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["data", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "00e30e7ce5e4dd90566968c774eafce3e66f6205"} {"id": "train_045059", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["coding", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "5795d7df9ac6c8fb5b4bf8e51d86fef7c90eb5c7"} {"id": "train_045060", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 9,\n \"confidence\": 0.9,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "b1aebfafbdb5f42505fa087959c9aebf49580533"} {"id": "train_045061", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["governance", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "17ede8f9c8c3a8601e26b9337a79b15bc7f3e102"} {"id": "train_045062", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "1bb8d723c0df1c9ee401b31571c72725b2ea72cb"} {"id": "train_045063", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["agent", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "e9c15abc9eefe55c266945f616f50d76961ec359"} {"id": "train_045064", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "dc5a570bc5f2ed26b89ab9af0a132ff49d773508"} {"id": "train_045065", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["governance", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "9ee08b3c4c64ecef985eaf02947457cf9823e6ee"} {"id": "train_045066", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "24400ad10e1e150fa6e7261bfac5018a582b6e61"} {"id": "train_045067", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "576f543c39fc1d49a2ec1cc8c4ee227cd1af73c9"} {"id": "train_045068", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["security", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "29e46f9a233e53b81e021d6809056d1e58994734"} {"id": "train_045069", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "8d360de7f18d067ea53feb326a5a62a4988d5fc9"} {"id": "train_045070", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["security", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "20e58e89656b90474d6b94d7b8088bc025867832"} {"id": "train_045071", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["research", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "c0b084c12c1637a6c5b3352074dbec491b62192e"} {"id": "train_045072", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "48da2302a2b24d2e8f827af8612456dbf9ba5564"} {"id": "train_045073", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "452c2c6c6742e1e0b45ebab6720c1938b784299e"} {"id": "train_045074", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["data", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "dbb56d174263954747cf273240477180965059a5"} {"id": "train_045075", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["research", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "4e75d8a1027f37af352fa56204cbaeeb81a98393"} {"id": "train_045076", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["governance", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "f770078db62fb8c3e479c469096faa78dee46888"} {"id": "train_045077", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["performance", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "16e021674c2c8cba2806550e2ae2eba7acdeaad9"} {"id": "train_045078", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Python", "moe_experts": ["governance", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "cbdf9b06be1d0eb7ea3c9243e9dc68e7bc4530d9"} {"id": "train_045079", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "1ebe4844e0397ca4b2123e8381b3192e4600496b"} {"id": "train_045080", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "SQL", "moe_experts": ["coding", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "5850d877e97adc2f71c3aa26fcb5c6e0430c33ca"} {"id": "train_045081", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "7c3419b16f1c1545d0e04c7792b193ddbc9a0ac8"} {"id": "train_045082", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "1bba0756a3d4bd69e88a70d01f11eaa61ae8f274"} {"id": "train_045083", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "6c2c650b51fcb18419e25d2715bb4c415f28b8df"} {"id": "train_045084", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "8fc8d6616db2578eccb103b58a88af1c0c035e1b"} {"id": "train_045085", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "0340317aae39299daa5be761fc0ab926fbd1dcfb"} {"id": "train_045086", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["security", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "a5fb58669cc0bf696456702a6761bc1a94d51682"} {"id": "train_045087", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "6b756ecdadccfd91a4b1d462a222ecf42d8cf358"} {"id": "train_045088", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "840e943b126ca894fd9305b78e812ead720074c5"} {"id": "train_045089", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "6bfd73ca33a7ca9cd487096b7a51fa0099a3d882"} {"id": "train_045090", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "839a1c68f3962f767956c35bd9998391fffff1af"} {"id": "train_045091", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "061fcfe72aee4f01f845f0ab731895b0253846b2"} {"id": "train_045092", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "616abb073df88ae0d57a9735d64cac98ae7558fd"} {"id": "train_045093", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["security", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "0a4d1c125c9fd25f72cd7d5db32a91cb4b84a895"} {"id": "train_045094", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["evaluation", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "f36105498e31936f17a1f3b6ba47543c42bce675"} {"id": "train_045095", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["governance", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "c9d27069241e536bbc1ac6754dc0d3ec74d8ce4a"} {"id": "train_045096", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "efa3a9a020da16f72f5f7a31df20d676c8c63f29"} {"id": "train_045097", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "8a1a5c546ec2a4f5ee27fdd46d37e39eee2d9e9f"} {"id": "train_045098", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "555102b4c54b7193bf1ef943846e2ce693e10ffa"} {"id": "train_045099", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["coding", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "ec2418059dbc1dd82e5bda6e7c0b303844204b8d"} {"id": "train_045100", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["research", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "539efebcf19c424eeb67a9fc614083ab2f30c146"} {"id": "train_045101", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["data", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "f1b160bd15f08ffa3328092e5776f31afba73d12"} {"id": "train_045102", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["research", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "748cc9644ef5615396c4b9800198ef75f7cf2c06"} {"id": "train_045103", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["research", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "901b76c975e24aa94b91bb334835d4c456cdfaee"} {"id": "train_045104", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "26651d3078515750d2155b3deab076bff65f330f"} {"id": "train_045105", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "d0e940910012aa13d609338db9af2c2686e08dd8"} {"id": "train_045106", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["evaluation", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "a3ecedc758aaf938aa3791d876a2499f4e5dbdb1"} {"id": "train_045107", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["evaluation", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "4a603d96453e90f568fc3d570c83b87be6665adc"} {"id": "train_045108", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["governance", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "b288bb79cd1958f250c3e21633e472de7f9249d0"} {"id": "train_045109", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "f8eb1851ab5e31e3be35f4dfe95bf9f942c9c786"} {"id": "train_045110", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "fe7d06e4fc885ba9c63a284f258f5ef24779c4e1"} {"id": "train_045111", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "SQL", "moe_experts": ["security", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "13d3d3cad7dee390a5eede8b5dd1a94112a4d9d7"} {"id": "train_045112", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "SQL", "moe_experts": ["coding", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "d97db548be99a559938a715f32d03f4a0312e204"} {"id": "train_045113", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "ace0459307faf1c7933f1bc61a4d8bebfec67088"} {"id": "train_045114", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["agent", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "588294c820470349cc13e91be0b756f5bf782093"} {"id": "train_045115", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "b88e27a3d2e44a7bd7d97e381f438687578dbef7"} {"id": "train_045116", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["coding", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "baa9af1e9148a9d2d8805b3d7e7b3f503eb17e84"} {"id": "train_045117", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "643b6fc717244e138a6492f76a7c1fb70268ad53"} {"id": "train_045118", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "7ebe83fa77e4ba1b4469232923fb26685a886a61"} {"id": "train_045119", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "4fa77f57082769092593e5780d27dae462a0502c"} {"id": "train_045120", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "f7cc7c00df0f6a3c9087f632a722c89203b2cb35"} {"id": "train_045121", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n cost = 0\n for _ in range(budget):\n p = plan(issue)\n patch = edit(issue, p)\n ok, _ = test(patch)\n cost += 1\n if ok:\n return patch, cost\n raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["research", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "35aa639c03ed1633860bafaca052a0b9b984e663"} {"id": "train_045122", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["performance", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "2ebac9e136a595748fc8c8852a941f0f65c694f8"} {"id": "train_045123", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "8b534d5473e61a03607686bc9c44888edb4185d8"} {"id": "train_045124", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["data", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "a1b84307dfd8190016c0205c863736493f46bbd3"} {"id": "train_045125", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "af221d5529e66af03d2a4383d0891cb920c949b2"} {"id": "train_045126", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "daffe6544cd5b9d65b815e9cfd1a997410b57289"} {"id": "train_045127", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["agent", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "d1278d8e4b8b5e0edfd54d07889129020962a1e4"} {"id": "train_045128", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "519b58e6318c49f315018848bfffa4184c194f4a"} {"id": "train_045129", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "c23570af4257c1e858da2826bccc828810653731"} {"id": "train_045130", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "c337a1555ed8b0796ccb49997b2b9a0e18e7c96b"} {"id": "train_045131", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Java", "moe_experts": ["agent", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "08d9650a43b25b8629b2cc7dc459bec5123d66f0"} {"id": "train_045132", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "268340d7d58cc719ac7a25118690b4f652248789"} {"id": "train_045133", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["security", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "694db0a9bb0a5ee6da216836b1863a72d9196508"} {"id": "train_045134", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["coding", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "f34e6d9c53f59f2340b9f7ec3e136f708e15be5e"} {"id": "train_045135", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "SQL", "moe_experts": ["agent", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "c012639704990f535e8727240b5876451b403b38"} {"id": "train_045136", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["governance", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "081aa17079ec748eaf4f5ca708f9e536daea2ae2"} {"id": "train_045137", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["coding", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "748a6e819da525a75a09e24322ae425f1c8703cf"} {"id": "train_045138", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["evaluation", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "684c479b32990eba98d9d0d7ef82f35377988503"} {"id": "train_045139", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 9,\n \"confidence\": 0.9,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["research", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "b7c3ba471d1e25ae1620910fa7e9026d27c08589"} {"id": "train_045140", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "e038b87808c5417ab195b990112885b1eba5163e"} {"id": "train_045141", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "0984e349f86ea65042f6ea51aa9e95c434ddbfa6"} {"id": "train_045142", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "ff6bf48bb8c6a63444d0005e1823ccc599aff13e"} {"id": "train_045143", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 9,\n \"confidence\": 0.9,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["data", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "f21cd979230c205d4b3a847038cc94a950063b5f"} {"id": "train_045144", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "4dbdc64fab270ac9e1de9fda6c6f019bee295462"} {"id": "train_045145", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["governance", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "b0dbd6a60eaa03d9a9b4e2e1c5a27aaac0c1fb53"} {"id": "train_045146", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "5982e7a913887e1be9602c3c2754a7f377dc6108"} {"id": "train_045147", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["research", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "c6f0a02b94d6e947b80083ae8d34a69623650476"} {"id": "train_045148", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "b5349946c26868b0229b6336c2bb56cb9d6ad41d"} {"id": "train_045149", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["evaluation", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "734a96bfcbd42b24f6f27ae8971f20ea67f39cd9"} {"id": "train_045150", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "fd318346bcb3b193229cc16f19b50d65386f4c06"} {"id": "train_045151", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["security", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "c54b1c14a03b4c0d46c278d02eb814c6f4e02cc6"} {"id": "train_045152", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "7b622aac026be1f5b8bb4081ac71d78087b8245a"} {"id": "train_045153", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["agent", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "b1a6680afa307e064f7be1a8056b3d57b15d45d9"} {"id": "train_045154", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "6425658c6e042781d2f091e6be77d8dbe6b8d263"} {"id": "train_045155", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "d4c234d6015c538566ee98c7e315775eff4e3446"} {"id": "train_045156", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "eda0b3b7196989557bb87bb55c5413919173aacb"} {"id": "train_045157", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["agent", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "29b9a7b719fe6cb38ff480cbaf4d3742e34e2abd"} {"id": "train_045158", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "5b91297cd84ec7ecd71456ace2c326785f8f8f42"} {"id": "train_045159", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["evaluation", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "b2430c6a98647169bde25037242c9dbe5a799a04"} {"id": "train_045160", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["research", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "188e46e0fdbeff045c6841f7c80cb3b70662f113"} {"id": "train_045161", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["security", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "cdc67e0d70dd2325f61bad4a5ede4cb562fae3c7"} {"id": "train_045162", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["performance", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "3ca17d4cc877e5d81ee9b07c808cc3b04ce18fbd"} {"id": "train_045163", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "47aaec5058a77ee58f48b45ed51d4600956aa3d9"} {"id": "train_045164", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "C#", "moe_experts": ["security", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "5af898491fbc9688545c91d9b054e66fe4a90c0c"} {"id": "train_045165", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "f165a8a0099a6be3242f88305dffb16d67b80c79"} {"id": "train_045166", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Python", "moe_experts": ["performance", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "8c60bb6a907b93c8b7be7aa7373db9f9283fce27"} {"id": "train_045167", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 9,\n \"confidence\": 0.9,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["coding", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "8cc6d133cfc8a60d2c875833ec20e44a007afe44"} {"id": "train_045168", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["security", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "bcfe4ed7c23a8a5775a45dd1019cadae6845bd15"} {"id": "train_045169", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["data", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "cd2f0bde2cfe9a3c01ebcc0cfe4ff3387b3cfd7d"} {"id": "train_045170", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "a5d1817b7f84ecc034e39edf20376391bf28a641"} {"id": "train_045171", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "c15d49422f9621bdebeeaf6b147813e83cc004b0"} {"id": "train_045172", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["performance", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "4060515ce3fe49af9175488125003e5d37488991"} {"id": "train_045173", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Python", "moe_experts": ["performance", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "8c60bb6a907b93c8b7be7aa7373db9f9283fce27"} {"id": "train_045174", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["research", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "888420f6a746982494ff01503c5b86922bc038c8"} {"id": "train_045175", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Python", "moe_experts": ["performance", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "02b9005eff73bdf3facb37999aed2c260a018b41"} {"id": "train_045176", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["evaluation", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "28c70a8df8a3c0376fb794ade82d6ff0541db91f"} {"id": "train_045177", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["data", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "ee8589fd2e1a32a6a46ae3c931b2d7b5b583e09f"} {"id": "train_045178", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "68b154edde897c9b4d5efc7803846ae06f31e618"} {"id": "train_045179", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "02edf614681928a84bcac469560158a9c6121db6"} {"id": "train_045180", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["coding", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "458f14dcd883e04df43989012b0c43d2322795f1"} {"id": "train_045181", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "3eff1572354ea78aa37096a6825d7c488eedba0f"} {"id": "train_045182", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["security", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "cfc4db1b593f48c405d0df59d5cf838c2356d702"} {"id": "train_045183", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 9,\n \"confidence\": 0.9,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["data", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "6ec31760328445436cb8c7b7ca715bc0a4d0efa2"} {"id": "train_045184", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "2b0f9eb3635935454cfe4f6e7d0b1bc6e92b8a60"} {"id": "train_045185", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 9,\n \"confidence\": 0.9,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["performance", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "2ff832f76be904ebe87a2e9c0df34018624ec313"} {"id": "train_045186", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["agent", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "b42134a88b87bae7e85ce1c3a545a05f3d9385be"} {"id": "train_045187", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["security", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "7f0806dd59fd55133743d227aa0b229211927e01"} {"id": "train_045188", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "ece07dc8edb94cad198d6df36d8ab30f4b7f56f1"} {"id": "train_045189", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["governance", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "0fe5ea2bfd49d0f264fdc3492d5e24cc077ee5ff"} {"id": "train_045190", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["agent", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "426dc30f1c4460139ad4a154a4681b4d99f0b1e4"} {"id": "train_045191", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "d95d3dfa14798586f1207f103be0f186a41d7ca5"} {"id": "train_045192", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 9,\n \"confidence\": 0.9,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["governance", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "2074d1b96d9d6fc7dc27902720b68ddf1fceae0b"} {"id": "train_045193", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "ca26afba3d1237dd1dc20ac0b5ec44b4d38ae1fa"} {"id": "train_045194", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "7f7fded172f2fb4c97b4db36b95cfe9c0cea9b88"} {"id": "train_045195", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["agent", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "e03e736ab915c07a2d7b6415cb7369d8fde8c79e"} {"id": "train_045196", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["coding", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "d0cb3c1f2ed1f61686feee91342dc1487718daa4"} {"id": "train_045197", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["data", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "f8ca4421cec1b2ce6ab4bb93df21929b7b8f005d"} {"id": "train_045198", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "e0d1bf5c90a5512128bbd45d4764ee3dbdf36e8e"} {"id": "train_045199", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "21ba2f689a4b6b14408f3828a24521c16885fb23"} {"id": "train_045200", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["research", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "024f90b6a3aaa8cac3b99284bf6cb75cf12ee399"} {"id": "train_045201", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["security", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "6596cce51b075b3c7e0e51b364d955261b2c04a5"} {"id": "train_045202", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "ba2e3c573dfebf0574ab200f5900100d979fd3e8"} {"id": "train_045203", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["performance", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "5e0f4231d1d2232a6c1f8ac70017fe4ca40700f8"} {"id": "train_045204", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "c095edccfd191e392c70ec4fe01f75825f2e594c"} {"id": "train_045205", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["research", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "ae2f83e0a453f3ea93710a5cc384257cbc61cff6"} {"id": "train_045206", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["security", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "9edea02502e060fdbda2f477ffd16a1c30d0e637"} {"id": "train_045207", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["security", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "f828b7abfa2328107b1152013d5710d5140e0d2c"} {"id": "train_045208", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["agent", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "53085934620a280679b11e8c51797ad64b0a2cbb"} {"id": "train_045209", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "9be64a00c73c98aed0dfbf9ced46bd37c98bc9ae"} {"id": "train_045210", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["security", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "de85f8dab35be14d0b5e2fe728b6891ae1370e5c"} {"id": "train_045211", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "4c4b66914abfdd732bcf425a0af1021cba40520c"} {"id": "train_045212", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["governance", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "944c2efe664129d61b2519bc5cc9f0dbc7c20166"} {"id": "train_045213", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "c6027552d83858dd21b5e01500ef4ab7b5ce5635"} {"id": "train_045214", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["research", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "173351ec0b8ca3cea781fdddce09fb73aba7a160"} {"id": "train_045215", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "53a985d4a3a7dcf7bc6276e4653386f02356e625"} {"id": "train_045216", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "e17e15f608f063060b39e438e48abafdcfa5c693"} {"id": "train_045217", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Java", "moe_experts": ["coding", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "c14ab374b8584a723ee18b082e52b7a6cdb91a2d"} {"id": "train_045218", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 9,\n \"confidence\": 0.9,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["security", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "2da75113c2392a99dee42fc79fb061557f51f759"} {"id": "train_045219", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "cec172dfa3dc12f0136923e5bac01b15030830af"} {"id": "train_045220", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["security", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "35d957bc34704b1d4ce1ea6dbcee644964dca3c9"} {"id": "train_045221", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "b89c3911ca8244619cd016cc0585f9b1742cd7f2"} {"id": "train_045222", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Bash", "moe_experts": ["agent", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "ae5f0114add3e7f4698a1a0f0351f1a52261b5a6"} {"id": "train_045223", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["security", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "aa1fa62b4db1a0ac818de8ace5ad33660bcdd111"} {"id": "train_045224", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["research", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "96b3205d874daec2b79ac8ca23c048eb061b6321"} {"id": "train_045225", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "ba5f76daa1faab8e7ce623e81232b66b24f6d7d7"} {"id": "train_045226", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["coding", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "4862643bb76e797bf7d0e507d3a1bbf30cae121a"} {"id": "train_045227", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "af9fb3c897591526de2fba8eedfd64dcf75479e4"} {"id": "train_045228", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "5aa272da6c06ce0bf9562041bdcdae44661424fa"} {"id": "train_045229", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "8fd68f31a6b2064ead733a2845a7cf04c149d4a4"} {"id": "train_045230", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "9fdf79a16f3179d7b99647b17686671d6cfebf8d"} {"id": "train_045231", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "016ee965bf128ab58386bd98e14d6761f58de5e7"} {"id": "train_045232", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["evaluation", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "4862643bb76e797bf7d0e507d3a1bbf30cae121a"} {"id": "train_045233", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["security", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "78b99b1f3d028c65a99c672425c103295f49bb96"} {"id": "train_045234", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "ce9f4ed6fb088b9757e5d9a6a5f3889bf7ae0ac9"} {"id": "train_045235", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "09df9aa2aafd1a3a25113ef21b13601287011612"} {"id": "train_045236", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "e6432834eee4503727e946281b7efd786da711f4"} {"id": "train_045237", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["governance", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "fe424593c115104b1fd4d40fb5ac3da2aaca36fa"} {"id": "train_045238", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 9,\n \"confidence\": 0.9,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["agent", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "79614b7e60fbdca2a6234769b2e843609e20bbc6"} {"id": "train_045239", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "6f942639479c07786c5063cca45759dce8881505"} {"id": "train_045240", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["data", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "3973cd4996bbd084f427023e47b30230b52f7846"} {"id": "train_045241", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["evaluation", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "1ca4cdc8f6c576cc90db5e5ac5cf11f1f66ed530"} {"id": "train_045242", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["security", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "e2201fc88abf1ba41c99bef93d5f2b105fecebc3"} {"id": "train_045243", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n cost = 0\n for _ in range(budget):\n p = plan(issue)\n patch = edit(issue, p)\n ok, _ = test(patch)\n cost += 1\n if ok:\n return patch, cost\n raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["performance", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "6b5a4df28f11f41248129d993e2083da5c52ade2"} {"id": "train_045244", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "d38abc67c45637753618d39f7c7462020a59083f"} {"id": "train_045245", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["coding", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "6a3c280e74bfb1f4c8a14dde4cceca5b6eb8e920"} {"id": "train_045246", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "1b94df5cf69d341fc66f42d3988822190a4d943b"} {"id": "train_045247", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "84e26c86b3fdc44a021ebcdbead889a42e20f8e5"} {"id": "train_045248", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "1ffc0b99889a2cbe808fd25582995e68b5ba2b7c"} {"id": "train_045249", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["security", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "2a2b92cf94a364801c4417f52c3047b6e2ce358a"} {"id": "train_045250", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 9,\n \"confidence\": 0.9,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["coding", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "6b886d6bfb34b3c7c182825b9d132139be6c0075"} {"id": "train_045251", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["data", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "4c7e01f21b1f340a4616b7ff27735c25de82fa6d"} {"id": "train_045252", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["coding", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "cc41f2b517d3d56b5470e56076c00e9cb6ca9ce2"} {"id": "train_045253", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "8f261c80109e98f0a89799f189fea8ab0581a22d"} {"id": "train_045254", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["research", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "a48657be5ed48805fa41de4e14cffc423a785cab"} {"id": "train_045255", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "6a22999336b64388358e41faa20b0e78d2646d7a"} {"id": "train_045256", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["research", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "7b25e8f5a65c7fa5d457e8a6809dddd9d96a6355"} {"id": "train_045257", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["research", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "858a45dc996ddfe067ce3d3a19fc748c013a40f1"} {"id": "train_045258", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["data", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "f082631a5bc6b544a1e6c8b530ba528b8f198018"} {"id": "train_045259", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "9fcc94d1f09f2223a5982387dbba473c85f88fc9"} {"id": "train_045260", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "29429b1a1662532460d89455ca9a34b930be2378"} {"id": "train_045261", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Rust", "moe_experts": ["research", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "3c37768fd35fd837d816ff4c0be25f10911d39d6"} {"id": "train_045262", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "92e189cfc14d8bdb02e0c3f26a7a21e42ac1abff"} {"id": "train_045263", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "cff88fe6be4de375e8be7239967106fa61966f41"} {"id": "train_045264", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "828950b478927cb1238c7a0448fe7570fae1332d"} {"id": "train_045265", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["coding", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "c44d2a3609e7b3b27ee3e6cd96caa76f975a8136"} {"id": "train_045266", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["agent", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "cb8c6c9d3403ad6fa514319d66b856f54216874e"} {"id": "train_045267", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["agent", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "ea2d2c0895d8a214eeb2675666850f909797a4a8"} {"id": "train_045268", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["security", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "b76e049f4139d930bd238e0accd2cd457113fa5b"} {"id": "train_045269", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "b958b9bd0efd2b33528129db7a1f9cc1640a61d6"} {"id": "train_045270", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 9,\n \"confidence\": 0.9,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["governance", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "4de51e09f2b7640d4d86075c652715e170f73516"} {"id": "train_045271", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "d225b742e01776cd9e97b49b4a51f42e0061f357"} {"id": "train_045272", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "2ae6a3d2e9b285236b58591d45c310e9acb54879"} {"id": "train_045273", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "f68b3cff0afe292d404d38e4e0dcd2e35afe04e0"} {"id": "train_045274", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "4a2698860d9a89dff49d899f9529bdc95c56e5e8"} {"id": "train_045275", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "9af9562d7879b67adf2f1126ee1c7cb3e7de2c16"} {"id": "train_045276", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "5aed8cea893eac75d0add5c2dd3b7e0b802db8a8"} {"id": "train_045277", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["research", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "af3978b6506f351db3d3bfd38d3c0ef4cc2c410c"} {"id": "train_045278", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "c794ca3fa90f1add0789a83aaf5e8f7225eee5d5"} {"id": "train_045279", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["performance", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "3862ddef5dc8ce80c36ed4d9122dc54c7a02ab1b"} {"id": "train_045280", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "C#", "moe_experts": ["security", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "3f3f92162e1f786bf4743186d80fa9527f0856ee"} {"id": "train_045281", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["coding", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "94f64a7643cef783d060b8928c32ab64bdc00006"} {"id": "train_045282", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "06ac70315bc28afb018a13d22cf40dcce0ce8ade"} {"id": "train_045283", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["agent", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "c7957509b38ce658cacd7b84a1cdbb0e22900407"} {"id": "train_045284", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["security", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "286da499588c57754f6955b64daced291b9bd719"} {"id": "train_045285", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["performance", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "59a5f0a035a75990af00535dcf976da513eca136"} {"id": "train_045286", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["performance", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "4beada11c2131682bf2e0e32853b4f5e8eea8a80"} {"id": "train_045287", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "1d69de07268ecb130fae928fc921c62d4e92ac62"} {"id": "train_045288", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "72b3f4e1b2485c525f3f2e1f2c2899326eea8fbf"} {"id": "train_045289", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "092485e0c3145d8a3b61101511b6c68c20f99848"} {"id": "train_045290", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["coding", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "2bdfc117167b3a1c40b5e2737ee29ec28560a93f"} {"id": "train_045291", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "74e628c4638502ac8c5794c71b3a2aacd7968622"} {"id": "train_045292", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "64eb9aab65faca32eb8fb480ba666f06b7d12f39"} {"id": "train_045293", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "939c166a1e4eff1bda26f894bd68a5fbba23dc5d"} {"id": "train_045294", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "72b049613643ea818690d53cf896db929e38b290"} {"id": "train_045295", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "79d25f3b7d3d61d327eb8ba0d1adc0f77a2e8127"} {"id": "train_045296", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["coding", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "9b89361a0946fc65d93f0e23fc791589f918ec16"} {"id": "train_045297", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "1f73e11930b66938992ff0c12d2abdf6cca9d5cf"} {"id": "train_045298", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "edfdfbff4a7977f5ffbd5e340a3ef6128b06b218"} {"id": "train_045299", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["data", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "993d637c3433c6a1ad9964551d53c2eff1380ce4"} {"id": "train_045300", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["security", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "ffe8dd698941844c01298797fac3ca73fe14a305"} {"id": "train_045301", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "370d407439060fd327994afbfcbe6b77bca140ab"} {"id": "train_045302", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "cb705017584ef1217fff07034b10efed623c3810"} {"id": "train_045303", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["data", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "a62f180fc6582936ea4d30c6ebf48a9c0629bb68"} {"id": "train_045304", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "eadc287ca81c5c0962fe214cbb2a42412c16e64d"} {"id": "train_045305", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["data", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "646735f7dc0db4ac01ea3cc12b9c2b25a9a1b4f1"} {"id": "train_045306", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "bc2615067ad98920d7b99055f699fa989ec96dba"} {"id": "train_045307", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["governance", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "4b1c8c0e9769ce730685e0c383603df9c92ed9e8"} {"id": "train_045308", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "c664fd32416037e33a135585cb20ad57b525ccae"} {"id": "train_045309", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["agent", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "da1cea7333f2eb5d41ff3cc69b1e9a97a579bfb9"} {"id": "train_045310", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "3f95c6cfd2650fe1dd171aeef31ac3f106c1d2ec"} {"id": "train_045311", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Rust", "moe_experts": ["data", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "083d8f0662b2468da0a3dfbeca43663a880f4352"} {"id": "train_045312", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "50d48f81befd0f12e005aee1eb93af178d79efba"} {"id": "train_045313", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "a914a3a3b8ea306b53f7a1bd9f6f8a56f4eeb6b0"} {"id": "train_045314", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["governance", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "eb3b48476ee518be3a79a55a1c1770e3d99deea1"} {"id": "train_045315", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "f6807152e62e6e511a9152b928a1f2cbb3dec0f6"} {"id": "train_045316", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "344e635a4298aaed74d28f308ae9731e4071f276"} {"id": "train_045317", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["coding", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "764658f543ac8bfbc88cc7d8637a3089c29b563d"} {"id": "train_045318", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "ea54782a51569b610176734ace0ef4f526e850b2"} {"id": "train_045319", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "b5160c9e7859a5eb4e81470e32b9ea6ae64158c9"} {"id": "train_045320", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["research", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "f8296a93eb3aeda83a9504f0cf1619e37f9820a8"} {"id": "train_045321", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["coding", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "8885c729d85cb09de9e5d967635847e53d45175e"} {"id": "train_045322", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "7d56f7a38108bf65398218c8e4a92c376491dafc"} {"id": "train_045323", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "72682fb9057eb9cd6615c465502eeb613810710c"} {"id": "train_045324", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["governance", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "3973cd4996bbd084f427023e47b30230b52f7846"} {"id": "train_045325", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "C#", "moe_experts": ["coding", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "c3eebf54a00725e8cfc14e0cfa21d42c0ae5347c"} {"id": "train_045326", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["governance", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "9eb1a449a8f0bf77f47c53fe939572a89ab9e6f5"} {"id": "train_045327", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "0eeebcdfac311dcf9f253ab5706f4294ad2911e6"} {"id": "train_045328", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "2325faf740603a82df066d181fb1a314308c662e"} {"id": "train_045329", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["security", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "6ee143e19c5c85a24104ad4e6a043aea18bcc819"} {"id": "train_045330", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "4cb2d692e0ce6b5b277d42ecb091a5facf172bc0"} {"id": "train_045331", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["governance", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "bdd778786ee22fe0abb659a4c062ef9091d1a1d3"} {"id": "train_045332", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["agent", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "5aa9ab865ba54b52c68de2b63975146e24700a6d"} {"id": "train_045333", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "8ca0f0e9e46eabb3e6960263e7c1015253c78ba9"} {"id": "train_045334", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["performance", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "800861a4aaa542faac69daaa30c735bc5e351c32"} {"id": "train_045335", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["evaluation", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "ec4791bf1adfdea36d4d9557489676082387e6d4"} {"id": "train_045336", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["research", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "efcd78c1ac745a88be55613d1e032358929c48b8"} {"id": "train_045337", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Python", "moe_experts": ["agent", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "b6d19ca66b381aa0c5ac0d7f7a6f6560397edcb2"} {"id": "train_045338", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "917816ed23a0b8d1ca6a9eb8eaa037a4e68aa81e"} {"id": "train_045339", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["coding", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "61d4c6951d2a7eae5552a6f697ca729a115b7d7e"} {"id": "train_045340", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["agent", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "33f55b1e963c0e3a78b42ab0a757e0d0eca9ca61"} {"id": "train_045341", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["agent", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "26dbafae46cf979bf1d27fed5033c9a7b0dc8e7b"} {"id": "train_045342", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["agent", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "8354627c6b78ecb8167c3ae55c2e6113aa1d5b94"} {"id": "train_045343", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["agent", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "76980d4128f4fd5c05c7886b30472dfcb2641e19"} {"id": "train_045344", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "a9cdd1b86c416a8460548c1cdc220827cc7aace2"} {"id": "train_045345", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "052e30ed22a5decff3cec73cd8502265a09a8dec"} {"id": "train_045346", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "7eac6b4c117ba4b3d6163547ecd115ba4a7318fd"} {"id": "train_045347", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["data", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "142a9469ecfabd2488398f1ca013300abe73d502"} {"id": "train_045348", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["agent", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "7fa07c9199bf56da5abb0da596a7179fabcde75b"} {"id": "train_045349", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["performance", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "99ecaf8be1f905af9f95aaaee0bd970fae9ab02a"} {"id": "train_045350", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 9,\n \"confidence\": 0.9,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["coding", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "8234a44df918ec598809fd3614aebf1106ad47f4"} {"id": "train_045351", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["evaluation", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "bdc5826598fa896af658672efeea68e21de27938"} {"id": "train_045352", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "e64747ca6d04d94ce0609e40a9d81d4f8755bc06"} {"id": "train_045353", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["security", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "9e76a4be50eadfe4a97a8e3f4d1ba38050a2c8ce"} {"id": "train_045354", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["performance", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "7f9648a8d3510366b1ca6467eb8f0f3dce763b3a"} {"id": "train_045355", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "6320ed841e49acb45d7c99bbdb031be2925d7520"} {"id": "train_045356", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "75403a37359dd79d988bb39b1eeaee2198b9a547"} {"id": "train_045357", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["research", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "9b355aa04352046e3a13e1043bf5f0cab2685788"} {"id": "train_045358", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["security", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "1a2f11cf614b2487b30737dcda03793ebf418181"} {"id": "train_045359", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "1dc491f79f71c52c0681b62ab2cf890175822521"} {"id": "train_045360", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 9,\n \"confidence\": 0.9,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["performance", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "faec4e1b622754f2cab429a957641e3d4f38ac21"} {"id": "train_045361", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["evaluation", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "225fae2d776e01af48350decfac733bcb97a8ebb"} {"id": "train_045362", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "e94517b47613cb5372ec10ac1c1531c32a7e720b"} {"id": "train_045363", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 9,\n \"confidence\": 0.9,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "f4e4cd53a5f5bd055fdfa0673519a6cba6c282e4"} {"id": "train_045364", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "b6148328ab4fcf577c80cbd8eb1aea9b4e89ee9d"} {"id": "train_045365", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["evaluation", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "2d2717510e415ff658e124280320b2cf4d6181a6"} {"id": "train_045366", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["data", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "1b02813710611dc2265d40fa3af447526a089577"} {"id": "train_045367", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["coding", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "6403cc058490676cc72f0c3cbe63d768d74f4c71"} {"id": "train_045368", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "7b0fc61f8ec5305f0e62dedff4fb7a170af3c2b7"} {"id": "train_045369", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["coding", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "85aa62301756b7940eb650bcd43ea521a85171e5"} {"id": "train_045370", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["evaluation", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "d20b375672691e6f4f85e1fcb068f8823206815b"} {"id": "train_045371", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Java", "moe_experts": ["governance", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "e5695c329fddcd4395c223d87ffb3b445f0b20f2"} {"id": "train_045372", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["governance", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "2731b1cd75a9b80d3f4cac06ce90f99d7616b01e"} {"id": "train_045373", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "534b4df7a648e1162342e164439c5288f53224d4"} {"id": "train_045374", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Python", "moe_experts": ["coding", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "d82aad6c2062bef104443a454f96557d9a359f75"} {"id": "train_045375", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "90ccd49a2bd900e9ad8757afb07fb99b50d4364c"} {"id": "train_045376", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "cc2a006289a115c40dd8acbc9a77a19b0c8ac63f"} {"id": "train_045377", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "b1452502c9144a138b98fc85d355ae30da019a7f"} {"id": "train_045378", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "7aeaf9e9fafb5b1d883926d80b060e410447a6f7"} {"id": "train_045379", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["coding", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "14100f36a452d1249ecb95760e2614388d3df374"} {"id": "train_045380", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["research", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "8bc3c851b85b971056e28ea9ca9336b73890fd20"} {"id": "train_045381", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["data", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "db0db371132cbbef8aaa0fa0c91529d5e1f11363"} {"id": "train_045382", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "156147a93484a2ffd9cce98a6bfce1ee816762fc"} {"id": "train_045383", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "1f651ab450d842a23a08a3f25a7abd9c36721d41"} {"id": "train_045384", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "6232a9557517df5d435b6eb6c06cd1093edad9cf"} {"id": "train_045385", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["security", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "09ed7f0b5a1124528ce717de30ffb93e9edb053c"} {"id": "train_045386", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["data", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "f529fb948558e0bdc6fd031f9232fd1e808e4db9"} {"id": "train_045387", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "eb409a4e0c4f9dd6e6f3aff3184b20426029c364"} {"id": "train_045388", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["coding", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "1dfd0ec7977113c3fa45c1052fea170dfbe71ad6"} {"id": "train_045389", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "7d56f7a38108bf65398218c8e4a92c376491dafc"} {"id": "train_045390", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["data", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "7ea013882609b487ac228d7c6f8fae4aba2b0964"} {"id": "train_045391", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["security", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "b67c6d9a046c81ac956a829a2e72347599a229be"} {"id": "train_045392", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "49f48e72b8cb895826b159a63061c8dcbdee0996"} {"id": "train_045393", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "cc3a9388eb540f2592024697a895f68c6e4cdbb7"} {"id": "train_045394", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "c4ed4c713b395abf8478f98799e9c1ae123e36af"} {"id": "train_045395", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["security", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "853969ca7fa1a28e2b1b53cdf8928db57d174161"} {"id": "train_045396", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["performance", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "2592dea66c6bcd86fb858c21dad5f705681f0439"} {"id": "train_045397", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["agent", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "7a7782c7c07d61d600e7b082cf93547a621090dd"} {"id": "train_045398", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["research", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "a2e529c85d6111c537bea9116449d382a4953f28"} {"id": "train_045399", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["security", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "d1030725953ee432a1643bd6f4adefbc9a594fe0"} {"id": "train_045400", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "b25ce8833636bd9f62fb1d17b68576297ecbce46"} {"id": "train_045401", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["data", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "cf5a2cc375594a1bf71436e793095c5c0524f9b8"} {"id": "train_045402", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["evaluation", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "a70931daf8b7021f0f0f121741cbac0275ff8dd9"} {"id": "train_045403", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["data", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "78c7202ac852362798c14030a0131420f134580a"} {"id": "train_045404", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "e107b0bf5f27498d9fe99aa6b420e6078c679008"} {"id": "train_045405", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n cost = 0\n for _ in range(budget):\n p = plan(issue)\n patch = edit(issue, p)\n ok, _ = test(patch)\n cost += 1\n if ok:\n return patch, cost\n raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["data", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "bd1bdc736389d468077cda22988808b0894dabed"} {"id": "train_045406", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["evaluation", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "d2bcb0eb45f859b812c2d0df6f640e8225b772cc"} {"id": "train_045407", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "1def973fdf19d624e61e6c2968812bcd79bfdb64"} {"id": "train_045408", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["data", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "a17cb0ed04bce1cd7eb9b74d9a1f371a791c5e5a"} {"id": "train_045409", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["research", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "57b700273ea7171d60c8e6b5897a63af0e7b18c8"} {"id": "train_045410", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "0b8038a058ba082e1e1f3eeb015b621ecbab0083"} {"id": "train_045411", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["performance", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "b06432f7327c839d525a23be90a57509ba8d3aed"} {"id": "train_045412", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["research", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "f0bcb6c15fc901eea6b4493b53c67c3ae6ab9421"} {"id": "train_045413", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "ae255d60483c6f2b300df323f80d318a62f99a2e"} {"id": "train_045414", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["agent", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "6d4bf5c087fed54945452df09c8b8fc56ec3dc96"} {"id": "train_045415", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "e0a689107e140b24b976d4558dc8c989dad44287"} {"id": "train_045416", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Java", "moe_experts": ["coding", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "943fc3e3ada7577de313bef03f5ea82caafe8fdc"} {"id": "train_045417", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["performance", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "461ecb93f25ea137385fa5765a2ed906d69c23cf"} {"id": "train_045418", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Java", "moe_experts": ["agent", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "55fb712f9161f0f14398217db900a6fed37ba258"} {"id": "train_045419", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["research", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "df442272fe474865be5a8ce4f00ce37f2d7a2ba2"} {"id": "train_045420", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "04ddadbb06a1aec8f3f45aa063990aaef7ee0fff"} {"id": "train_045421", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["evaluation", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "7df441dec679af77bb33a2b843157e3f2e72ba99"} {"id": "train_045422", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["performance", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "d3eec790d19022d14ca6b422b303aae5c2a86f15"} {"id": "train_045423", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["agent", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "6d17c50e66babffb8f9958013db2f2953477d52e"} {"id": "train_045424", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "ac6a0a5ab593a4f24d8c1af61842547815a3992a"} {"id": "train_045425", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["governance", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "4f4b80f1a0ae9bc17148b9f4b97593401ad24519"} {"id": "train_045426", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["agent", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "b7bce14fb4c89b7222598f0b7771a167ce595606"} {"id": "train_045427", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "547c9fa94c62549c99057589ff38af276f359600"} {"id": "train_045428", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["governance", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "0842ad3cb7479a412afa1b735ec7b686b20d65fd"} {"id": "train_045429", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["agent", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "2836e05c681eabb3942fa3371efc5e5de25dffcb"} {"id": "train_045430", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "c962a028f02a598da79ffa3dcc3178b446d0133a"} {"id": "train_045431", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "2e40cec487651e9843934a1b377f2c2198291b6e"} {"id": "train_045432", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["security", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "ac1a0bd1e374b9291ef3189c65d857c86fad1ca9"} {"id": "train_045433", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "09cb15711202218c2843e9bf623c904080c3369a"} {"id": "train_045434", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "76f0e2f5c2fcf88896b59572aba45bd6e77fde9d"} {"id": "train_045435", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "f34e7fa68a57e7eb286b35519d7625b43212ef81"} {"id": "train_045436", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["agent", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "c2427bf736afeb4f98eb1653b6c2f43a785ff2f9"} {"id": "train_045437", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "3369cf6323387ab9135f703fd5585a5c464c91e1"} {"id": "train_045438", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["agent", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "4d31982b72de03bf8259228cb0f55b7bbd77d399"} {"id": "train_045439", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["evaluation", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "6f355fcde0538fbf76bf9d7f97785db4474a4a57"} {"id": "train_045440", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["performance", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "24c2ccd5cfe97460623e91e63339f34e1575ecfd"} {"id": "train_045441", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "ffafb5ec0580ece919541c5c7225adb48eabc6fc"} {"id": "train_045442", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "471fdf4213764cb9fb52d3539bcfdf24bebad968"} {"id": "train_045443", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "be929c8da007a522a84f5f04f0f03e9cb8c48209"} {"id": "train_045444", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "5ff4b38a5cd2525b40d91e9bd81cbf9782d0dc86"} {"id": "train_045445", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["performance", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "e7ca2acb3ba0079feb35eb7b6d6fa697fa134cd7"} {"id": "train_045446", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["data", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "367d7ba120e55ad8363621d713303dc854465986"} {"id": "train_045447", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Java", "moe_experts": ["performance", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "0fefb3bcc91f4ce027301e98d29a9f2af0a57846"} {"id": "train_045448", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["performance", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "595d032ec62efd28dcd23ee7fc5446188903462f"} {"id": "train_045449", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Java", "moe_experts": ["coding", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "306b0345898e9249f7262c0b53d037803672c28e"} {"id": "train_045450", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "f6e624250a005d2bafb9af181c92a3e623e3169c"} {"id": "train_045451", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["agent", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "6f88730f0b476057a84461719224fe6f5337b59e"} {"id": "train_045452", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["governance", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "64a80556ec25737b2630ce0032c5d3ff564a2aba"} {"id": "train_045453", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["agent", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "9549f5a6735cc8fcdf83d7ef9f51bc693e653f59"} {"id": "train_045454", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["coding", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "fc49fa1152085e1f4c9c4f8b2dca0dfcad593052"} {"id": "train_045455", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Java", "moe_experts": ["security", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "961fd315fdecf0df4c892090539ec89646479f8f"} {"id": "train_045456", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "026eaab5afdde359f87ed0fc45776c3186192290"} {"id": "train_045457", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["performance", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "fb41125ff1c750bcdaac8d542a050d35ef384286"} {"id": "train_045458", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["security", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "030d9f74e16bf4fec0f057963f5ca770cd88fefe"} {"id": "train_045459", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["data", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "f2c20d2c67fcf4c14db4ef31acf1267c203caea3"} {"id": "train_045460", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["performance", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "d5a6dd06ea658b095eeb95db9d820017eaf69768"} {"id": "train_045461", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "3cc49d5e05d7418d2f3dc14c5132f3414a9af47b"} {"id": "train_045462", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "2f3e9561ba60ea80e089b0d0c5b46bf35fe2c189"} {"id": "train_045463", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["research", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "23a999a32725e1885ca34b968463f053bcae541d"} {"id": "train_045464", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "25c2400d091849be299ec096918c37c69a815514"} {"id": "train_045465", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["performance", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "44081405181e831e28712ed9197e8d6271b24bf9"} {"id": "train_045466", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "8fae73fd0ff02a6d0d7312b9e381913eeb10a79e"} {"id": "train_045467", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "0f93f9014deae671aaabdaa01d16bfe39db34688"} {"id": "train_045468", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "41747f8118ceb2ffec7d03adf66021078180f940"} {"id": "train_045469", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "76d8892282bccd04220b3527524748bff67758dc"} {"id": "train_045470", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["security", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "fa5b0f2b31dcdbaf8afe2a2b3931ebb09a4fcf4d"} {"id": "train_045471", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["performance", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "e394d9de662e36203fa06d382d12bae0238fded5"} {"id": "train_045472", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["data", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "8353afc2d168097ef5cb99ef0873f0e186f835c7"} {"id": "train_045473", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["coding", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "58c3ab781db2351ea813ba081b45b2aed2cfa525"} {"id": "train_045474", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "9a67cc5c6ed44687337fe2c006e9f2f7ddfbef14"} {"id": "train_045475", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "a9da9b27be60fd46dd5311861f0aa120a8253661"} {"id": "train_045476", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "7e9ac79c468ae04ea5c695e2945da60cf82df0de"} {"id": "train_045477", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "f0b65fee1a80c0d34b2ab0ed1c3d1c582fc98cfc"} {"id": "train_045478", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "7c35186fe7689275d857b2c2c4512e29c4d7153e"} {"id": "train_045479", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["security", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "a6b30229a4bb377a5c8a4dc02ea6b2cde0cfcbfa"} {"id": "train_045480", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "8096329737f7ef4155874a28a5d6e1a8a19fa90a"} {"id": "train_045481", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "54b81a2c06d878f70f79dc6238f9854fc0e5a735"} {"id": "train_045482", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "f2caea25253f9b2a8ef7397868652e8d1fc442af"} {"id": "train_045483", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["research", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "3d79b19abcce8441c1ce427b83e7fece379a0cd9"} {"id": "train_045484", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "b3a28948ac1c950c38eb82197275b141dcfc6be8"} {"id": "train_045485", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "47ce8ee96949f59c065b46d0cc0859598cb775fd"} {"id": "train_045486", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["evaluation", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "49d2154cb9f0d4a5796abb49002a2efb46266ee1"} {"id": "train_045487", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["coding", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "5b35aef15411ab36110703022c4672167f8e0198"} {"id": "train_045488", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["agent", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "bd2f242a614361cf88f5674e96c696753cfbd907"} {"id": "train_045489", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["evaluation", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "c8e3263567e3bf6deb08a8bf9fae6636209937ca"} {"id": "train_045490", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Python", "moe_experts": ["governance", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "9a7e08eb5f82e72101fa7e93203b06ec1600a832"} {"id": "train_045491", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["agent", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "6105d37e27e3ab211affba58609795f8e1c91b32"} {"id": "train_045492", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["governance", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "08d1ce89c2be3155fcee00d4f6150ae36cdb113a"} {"id": "train_045493", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["performance", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "c69d41ba412ec421889e3d2f6b68a4ab84441c00"} {"id": "train_045494", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "9c959812936dd0b622aa47821a2fb86559453071"} {"id": "train_045495", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "1c20916063afd738e4a11e7e970dfc41e47a40ab"} {"id": "train_045496", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["security", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "321e26ed591da28c2da72252feba4813d41ffe38"} {"id": "train_045497", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Go", "moe_experts": ["governance", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "311ec4643a3814992dd076900d8e28e85df57abd"} {"id": "train_045498", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "9b104a63e37b326c75315c911c319d57c53ba905"} {"id": "train_045499", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "312ed372eb2d3e79b1f03c382287ecf4abdc152f"} {"id": "train_045500", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["agent", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "bc319dcc138471c8dba646c25d7fce7c1f01218c"} {"id": "train_045501", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["data", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "af78777682407608382d401b1f65467158b64153"} {"id": "train_045502", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "d07bc29156fe0d916e3c340d33a24a6ea8cfba66"} {"id": "train_045503", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Python", "moe_experts": ["performance", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "3ed4669dc7fd1afe50e06a441e9b390a86ad32f6"} {"id": "train_045504", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Java", "moe_experts": ["agent", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "8121877b2b746c80f2605d3057b299ea0c8850dd"} {"id": "train_045505", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["data", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "2e842c9adfd0ff4fb35e4d60d45b2d11294f8499"} {"id": "train_045506", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["evaluation", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "015c91103adf5edaf11e56d620c344e9275d029f"} {"id": "train_045507", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "680ba1b60a2bc8a764056273849b0e7f94c40db2"} {"id": "train_045508", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["data", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "f2cd6b3d65da1eaaf8c6446e072a20cdc3bfb7e4"} {"id": "train_045509", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "3909eda411a808a15b316b322daa6c6709feae4f"} {"id": "train_045510", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["research", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "254180b573fceaeff88d1e7c944784f214653137"} {"id": "train_045511", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "ca980fc195d0ea84e1bed2aee74eb5c597f7f69a"} {"id": "train_045512", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "3816ce2136d336dfd7905932b8fcca3997bd4b4a"} {"id": "train_045513", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["governance", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "79361a4eb73570765088b2aeca6b61c5a3c1e071"} {"id": "train_045514", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["agent", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "ac5ed9855aa4d88d8d48c3b79c727ac48f224caf"} {"id": "train_045515", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["performance", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "db42af5a5e05a9b6699ad9641657f0c986b45bb2"} {"id": "train_045516", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["governance", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "cba47cdd7ea57a93948c5e229adf8fc9b847cafd"} {"id": "train_045517", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["data", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "c1a73e88451e84d35b3593e978af32b12dd62b25"} {"id": "train_045518", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["security", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "86c13d6169085eb1104bb64e4d79a04684696ddd"} {"id": "train_045519", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "a9fbfd5be32368b7343444e840cb7ac2240f658d"} {"id": "train_045520", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["coding", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "1080523e52e713f251af1b38f5fe5ba0387cdd94"} {"id": "train_045521", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["security", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "1c1024efbc7bfd5bf9e9c26da31069b68a9dd8a1"} {"id": "train_045522", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "e0e326255456e59c945e7f9f1b17bc0b11b5e3ba"} {"id": "train_045523", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["data", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "e2f97fe4338012755ef1d4ee0b52be89c49ef3a1"} {"id": "train_045524", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["coding", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "5ff28252d0c9e5eac8561e6ce8f4f2109eb0833c"} {"id": "train_045525", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "ef01cfb40d6c912ddbd0ab3f4a02d94155f48f59"} {"id": "train_045526", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Go", "moe_experts": ["security", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "5cb165728c3891c0959abca58b22efa2a733cdec"} {"id": "train_045527", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "1814a4724b63fad159dbcba5c14d14c7b5f2567e"} {"id": "train_045528", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "0b68cdc9bc1006ae914026914af6e42be2275b45"} {"id": "train_045529", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["research", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "187c367208e489acffbc283f13ba94b6b050efad"} {"id": "train_045530", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "a0bba63280841063c47f433dff0f42271ee9beb6"} {"id": "train_045531", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "3d0df50a9591ff3456a47d0afe3a845da4fa83db"} {"id": "train_045532", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["coding", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "12e78e160631b3ea2eb78a4ede0fa09d28d50d02"} {"id": "train_045533", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["performance", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "8c3e85483192848a5012933bb9847016e078f7a4"} {"id": "train_045534", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["data", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "8d783ac9e61bf36813e3618d77a289418327b976"} {"id": "train_045535", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "f5997184a50b927851701fd99f4936ae21553d6a"} {"id": "train_045536", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["data", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "c626434ab965d054f6f9cff7827ef026de840252"} {"id": "train_045537", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["research", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "fbf678bd4e86c841706d57cfdff8638f35dc105d"} {"id": "train_045538", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["security", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "0e7a830edf60a600b899c208ef84b2f58d80b290"} {"id": "train_045539", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["governance", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "6b20657054f768270c06a6a55339e440223ebf3a"} {"id": "train_045540", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "411e7bf8e0e91a33ef2664d99464769f274cabbb"} {"id": "train_045541", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["governance", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "944c2efe664129d61b2519bc5cc9f0dbc7c20166"} {"id": "train_045542", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Go", "moe_experts": ["security", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "311ec4643a3814992dd076900d8e28e85df57abd"} {"id": "train_045543", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["coding", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "b63d1faacd1a9c5bb0636bf8067db49a355916f0"} {"id": "train_045544", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Rust", "moe_experts": ["data", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "5ef24bb2509397538d056f4fe5b714cfb2a7b43b"} {"id": "train_045545", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["data", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "4f086522f1d7c9c9ec1bd9270861c8426a31b3ca"} {"id": "train_045546", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["agent", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "70f3133ac490354341e2f5f029cc13eee20bfe6d"} {"id": "train_045547", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "902618f2dd6899315a75e6be5774a447334d85be"} {"id": "train_045548", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["performance", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "1742755342dbcbaedd49f5d5d966e187f914e07b"} {"id": "train_045549", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["coding", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "33c5b6dd3286dc7838e74bc2c04ac7bd6fbccf31"} {"id": "train_045550", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["performance", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "02f53f4749e3c32140ef6901ac584c998ec766bf"} {"id": "train_045551", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "d58763abf56a580506f0cd7e7a82de4dad6e29ca"} {"id": "train_045552", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["coding", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "99b8a43ea48b39e0c648e3a65bbe67832d3f5054"} {"id": "train_045553", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["agent", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "861b73950b0fc596e25e0e6d34fef07e4a7894a1"} {"id": "train_045554", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["research", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "8dce11440f53cde53c91b0a80e8949484d57a211"} {"id": "train_045555", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "3a909aeabf8ea5cc4ee73223ae2fe222e57859df"} {"id": "train_045556", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["coding", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "4be1ea9b28831e02ba04281afc9a1b5474b92a5f"} {"id": "train_045557", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["research", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "9d5729b902318cf70d3bf12d4e4a1b2540ea5253"} {"id": "train_045558", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "ce5d647135b99d86136a9c57c3c9aa1510af2052"} {"id": "train_045559", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["governance", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "56c527bc170411ef7b0c30f020a2b23857ed2870"} {"id": "train_045560", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["governance", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "c94ff722de2b0ee10f3fde5ac2be9d38173ad649"} {"id": "train_045561", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "5cfa5fe63434deba762effafd2f25885481cb94f"} {"id": "train_045562", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["evaluation", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "26f68f4055fc0c19300e16e6e233d29d25525a60"} {"id": "train_045563", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "db6732db5be798f80a5d5f014564dec50c147fd6"} {"id": "train_045564", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["data", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "93f2ec280f4858e2f2f3da1b6edc4ac604a31a02"} {"id": "train_045565", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "474b703dd7fa59f185189e5d69aefb21df3b4802"} {"id": "train_045566", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["coding", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "bfd5fb80213496923073ef4a1f0289ab63bb8326"} {"id": "train_045567", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["governance", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "4421186002c71102f4e7e9c20d01712d21f5f19e"} {"id": "train_045568", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "9523ec4a9cc6fe6abb415a6384171c93f4bdea1d"} {"id": "train_045569", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Rust", "moe_experts": ["data", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "4a0ac8667a9fa750e016eb4b870aaad61be1987f"} {"id": "train_045570", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "31ee8fed58856d237e597edc990f69ea0afd21ba"} {"id": "train_045571", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "2bb1898dbab77597103f2054181361a149dba176"} {"id": "train_045572", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Python", "moe_experts": ["coding", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "d45bb2d5ba4dd3eec842589cdfe0cf7fa9dc7747"} {"id": "train_045573", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["security", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "3e9472a5e08651c923acc91cd7585b80c63dd116"} {"id": "train_045574", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 9,\n \"confidence\": 0.9,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "38432fb109ce660c2f6ec07a07b38675904d77fe"} {"id": "train_045575", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["research", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "c6c1e4b131c0896a31b5224abadccca2b6c11d1a"} {"id": "train_045576", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["research", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "fcd2aa462d81e2ede7bb8d2d47159fa0c0b5758d"} {"id": "train_045577", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "27bf7efb4f7e47435294fc3726a69aacf8de654c"} {"id": "train_045578", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "72cb5844a05ac23b91c6e6a216fa473f42c6b029"} {"id": "train_045579", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["data", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "fe2bd2e1ba29d6e619955de07be50574c5f5962f"} {"id": "train_045580", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "71c706e8cc2160fbe2856f3b75d4950e259ccfda"} {"id": "train_045581", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["evaluation", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "0ef3a293381b46d079005f84c1eec110908088a4"} {"id": "train_045582", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["evaluation", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "6b9a2a17a46ab25d2ce2ea14394df2327a19d434"} {"id": "train_045583", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "0f93f9014deae671aaabdaa01d16bfe39db34688"} {"id": "train_045584", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["evaluation", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "373a75cfa0004d5717ac6b2c96c3286d817bcb41"} {"id": "train_045585", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "72c3e5e37df11a2850d034fd6663e6059d00d26a"} {"id": "train_045586", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "a7135f87034ec0b24418e09ca8683dfdcb89a370"} {"id": "train_045587", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["governance", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "b303f75d0b67935bd6fb14c0f5c0fb0428367928"} {"id": "train_045588", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "bdf9027b16afe9b0a79cf2f72ca681fa2dc1c058"} {"id": "train_045589", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["coding", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "9165749821f61069b109518a4e2a5368cc67a612"} {"id": "train_045590", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "1814a4724b63fad159dbcba5c14d14c7b5f2567e"} {"id": "train_045591", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "f64dc1892e34f9171c2163849f677c50445c6159"} {"id": "train_045592", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["agent", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "ee063c1d358ec08d25aad03aabbf6609485460cf"} {"id": "train_045593", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "5b635b030f4a617a7d52f00b2bdd7650bc84b644"} {"id": "train_045594", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "9094fa3e3ef2bedb873fb96702afc27a844dee5f"} {"id": "train_045595", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "65f712ee270b488d213b3f8d0a74e66d1881cf52"} {"id": "train_045596", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["governance", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "986cb1fdecfdaad6d2b45850e0e7c431c5f9ded5"} {"id": "train_045597", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "678ae6324e32b2ea92e1005605a87384f785fa4e"} {"id": "train_045598", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["governance", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "f3f712ef3ff338a823ac1517a62aea4b2a168528"} {"id": "train_045599", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["governance", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "b58e7901b6e9ccdf331180dd4708ee9f2c970628"} {"id": "train_045600", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["evaluation", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "bc5fa6fe3bd46f49cc9209b95e977a8909d34393"} {"id": "train_045601", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "2117c4da758236600aae5bd37542b353887601bf"} {"id": "train_045602", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["data", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "61394271a259bc80384d0355e0c97c0f8ffe7230"} {"id": "train_045603", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["performance", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "dbd2804c059d2b55c3f582d7e100da957793b54e"} {"id": "train_045604", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "614662976a4d4d8333b340b4b38a5f94ffec3651"} {"id": "train_045605", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["data", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "7712c0949c6358dbec49e8284300684d541b73ff"} {"id": "train_045606", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "c2efb6a0d1d1c7f888908e0e02cb996c89dc29b1"} {"id": "train_045607", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "5bc762f153589aff60f5b8105c72ad7d98256507"} {"id": "train_045608", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["performance", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "73e799f8878046aae2fc476e212bf7e002143af1"} {"id": "train_045609", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "8550fc3199d752d6ab29849bdac9d85dbe250f1f"} {"id": "train_045610", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "aab8eda2415f049e65f2f08339e45cda998c35a0"} {"id": "train_045611", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["coding", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "4296753438debfe11172155ddf2041ded391b202"} {"id": "train_045612", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["agent", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "d357f3b1cd40fa1ebfc93e3434393645391f5797"} {"id": "train_045613", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Python", "moe_experts": ["research", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "ddd759fcd5207081ffc4f05798c17b54ddfd5e0d"} {"id": "train_045614", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["data", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "4ae32b39e316f24b98bd27f47d406b348ed9ff4a"} {"id": "train_045615", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Python", "moe_experts": ["research", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "b5ad848b3a467552dc6801d7ba840a5dd3ff90ae"} {"id": "train_045616", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["evaluation", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "41b6269636a0664fe56889acea58db49e5ea7fc2"} {"id": "train_045617", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["coding", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "f9a0404d6ccecd63d6a0304ea2c02b0173f57911"} {"id": "train_045618", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "c77a12e556039d53697a36f13299c44e0a1a4816"} {"id": "train_045619", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["agent", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "72c3e5e37df11a2850d034fd6663e6059d00d26a"} {"id": "train_045620", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "42463af8a7e73857de1b3d4e68b132ce25808573"} {"id": "train_045621", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["evaluation", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "b3d3907f94b00e42665980905da1de319a710dbf"} {"id": "train_045622", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "8be973f28c1812dc256c033f4f00727f659fadf9"} {"id": "train_045623", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["coding", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "191d09fe44beb77512481476dd6a32d2834771ef"} {"id": "train_045624", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["performance", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "e8190521c7d8a7ff6c4b2303d82d6fcc8715d37f"} {"id": "train_045625", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "988f55413c71b94e370219054209dc9c0e2647fc"} {"id": "train_045626", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["governance", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "95056e731df10a35dd23ba13d60205cecec2ac26"} {"id": "train_045627", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["security", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "22f89c2ce6b726398ff16721973f60491ca1fc18"} {"id": "train_045628", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["research", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "e89587a5c06f147ca6fedd17c8388cf780243441"} {"id": "train_045629", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["coding", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "c0b084c12c1637a6c5b3352074dbec491b62192e"} {"id": "train_045630", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["governance", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "f7b6aabf7ea65262445f8ac7417f90945e37f6da"} {"id": "train_045631", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["performance", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "e4128a92d2584c2aab50c9e1ecd2d270cbcb0c0a"} {"id": "train_045632", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Java", "moe_experts": ["evaluation", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "3f6b7a8227b96355831be5e0c3017bb2e04d4e86"} {"id": "train_045633", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["data", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "53fb9ab70d1c3fbcf32b88e4fb54e23c4d347df3"} {"id": "train_045634", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["security", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "badf5e120b9a1beb2cd5b3f7c9085182b31c2cb9"} {"id": "train_045635", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "f2e4ba6e1a48211d75369e9668d6d56ae31ec682"} {"id": "train_045636", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["governance", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "4adfb0c1ceca580e53e4f403b91577854d87c7fe"} {"id": "train_045637", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "90759ac86be26b6802a420206b9bc6880f60716a"} {"id": "train_045638", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "0bf2373ab01ce8c36a05f6c1b08f6c9683e95665"} {"id": "train_045639", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["coding", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "08ba4fb101460fd2c3ac39e3f921e9570833a0d7"} {"id": "train_045640", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "00048be3428c3adf6a84b4b83473656d3599cfbf"} {"id": "train_045641", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["governance", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "fbe02c19deb701725bb8b3060e8dd6353654d2e2"} {"id": "train_045642", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["security", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "94064bedd82711ef4fd31036a2061aff09c65c0d"} {"id": "train_045643", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Bash", "moe_experts": ["research", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "77b897b2259b7f7acf4560a38c8399b7d4030091"} {"id": "train_045644", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["data", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "1599889f21310b4fe8434041a4babf9a72238395"} {"id": "train_045645", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["security", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "8fc47c6e3b5dc6440ef69c9bc66073febdd110c7"} {"id": "train_045646", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["research", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "6ed8af993b8a1c606de091b362b354e1f551b66a"} {"id": "train_045647", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["governance", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "f58530a86c4a22bdfa559b7100823508ee8fd4e6"} {"id": "train_045648", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["governance", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "9efe89c01efc062c4f9da16bde9063ebd868ecbf"} {"id": "train_045649", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["data", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "99c25c2fd9951aae6eccc56b0a5d045c195cef9c"} {"id": "train_045650", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["research", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "8dd6650f3003bfc36d31e4081042889483f52468"} {"id": "train_045651", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Python", "moe_experts": ["performance", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "a239bf3b73dfe29b9e147115152a87fa9ecc1ddd"} {"id": "train_045652", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "d4ba4c39e7d907664f6d93d390c2ae07ba44b94e"} {"id": "train_045653", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["research", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "8e3c3a6f5e81ab7fa1514bf3c8f655eaf7854311"} {"id": "train_045654", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["performance", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "65bc4d00f6449d8b05962f8bd5b088a5a5d20aa2"} {"id": "train_045655", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["evaluation", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "e725c96b376c7af6956d9679968866f4f961f026"} {"id": "train_045656", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["research", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "06aa0ac0fdcd02a8f7d0dd1b4f467e23aab20aa6"} {"id": "train_045657", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "9d155b22dfdd2eec662fa600ffe7e6d485928799"} {"id": "train_045658", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "7a938905e360eb10f1249751f4722980bcfb7773"} {"id": "train_045659", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["security", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "f3e97f36bf66e896275a14db80c947c5c9af011f"} {"id": "train_045660", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["agent", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "69b12a3766eac5ab54b019bd585b13c26e3bc9d7"} {"id": "train_045661", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["agent", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "5aae1556d7b46f19f98f126bf26918f7df9281bf"} {"id": "train_045662", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "d0c2963ac3a060d8321f6a52158ab3b9fe4fbc4f"} {"id": "train_045663", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["research", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "8672a2cf4bf142912bb95e886d1a87dca8518ef2"} {"id": "train_045664", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["data", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "95b0a8af2df7b2ef359cb45d49c70fd4719a677d"} {"id": "train_045665", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["agent", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "23fda4cb48c34e2a4b50524501627add28a8079f"} {"id": "train_045666", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["data", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "8071a8086c7ac04ac27a595d01d1d0b00402f035"} {"id": "train_045667", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Java", "moe_experts": ["evaluation", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "1505bc9c477b6436533d778be7ea8a28e68f0a1f"} {"id": "train_045668", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "7c35186fe7689275d857b2c2c4512e29c4d7153e"} {"id": "train_045669", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "53c3acdaedf49ab1548c7c224975b24b81f98a7f"} {"id": "train_045670", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["agent", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "994e41a627ce2c8206a56cceb097d1b7412e9b6e"} {"id": "train_045671", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "f2bb02baf63497ccf1a715caf6e714d78f05b6a8"} {"id": "train_045672", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "85b397c2642b0169ad9600feb9722885993678ba"} {"id": "train_045673", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "832cc9da58606dc8637e276df40dde1f078593de"} {"id": "train_045674", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "820a23b9b658dbfacc13af6be34086c2801c4ccf"} {"id": "train_045675", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["governance", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "4aa3be3891417b456a0927ceea1034d7870ed272"} {"id": "train_045676", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 9,\n \"confidence\": 0.9,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["coding", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "b600fe45b894bb4b6f2351e78c52576dcb0aba32"} {"id": "train_045677", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "3c5e06167a339d5c020dc85f9e7f89b432237c1e"} {"id": "train_045678", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["governance", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "0afdaaaed6e45f233c5f65c5c159b3e106bdc24d"} {"id": "train_045679", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["agent", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "e7cae698e2bedf062a59dcd2b0654158adf2c19a"} {"id": "train_045680", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["research", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "20664917d694702cbac088e52e1cbeeb02534b1f"} {"id": "train_045681", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["data", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "60cc1c87b46fabf678e35e33afb66bdf31b34018"} {"id": "train_045682", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "5160a31db57211ea02af3e7c1f0249a1b2c3db1c"} {"id": "train_045683", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["performance", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "b58e7901b6e9ccdf331180dd4708ee9f2c970628"} {"id": "train_045684", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["security", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "485d970d886988a4a926cdf00196d059c4768aad"} {"id": "train_045685", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["performance", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "4e89cce436bf088e2befe1ed4aca8ae172d360ab"} {"id": "train_045686", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["research", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "7b44ae94b7258459932cb83ccf16cc787f6f1f11"} {"id": "train_045687", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "57de9c8205788c8369a5e0c40634421385a81fb9"} {"id": "train_045688", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "SQL", "moe_experts": ["coding", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "49672075a5085c49c49d277f565b0172dede9627"} {"id": "train_045689", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["security", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "e8b997c5f7e711e60b4bb1e7c0a2c4d7f9ba40f8"} {"id": "train_045690", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["agent", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "dc4bdfd6e750abde08ce8cb481feea5178a5992d"} {"id": "train_045691", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "2139c90611d733c52cadf412c509cdee1ced24f4"} {"id": "train_045692", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["evaluation", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "b3a28948ac1c950c38eb82197275b141dcfc6be8"} {"id": "train_045693", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["coding", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "37250226b4496443c4a411a3c25baf0838407707"} {"id": "train_045694", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "ec509c2d0a2f4eb91157fe5f23aadf996844c0f9"} {"id": "train_045695", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["security", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "ac1a0bd1e374b9291ef3189c65d857c86fad1ca9"} {"id": "train_045696", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n cost = 0\n for _ in range(budget):\n p = plan(issue)\n patch = edit(issue, p)\n ok, _ = test(patch)\n cost += 1\n if ok:\n return patch, cost\n raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["evaluation", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "b35497be8962717733b73533d4fcfc1a90b9ab9e"} {"id": "train_045697", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "C#", "moe_experts": ["data", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "5b263e10be01ef8bd6a58217f20233014b33584e"} {"id": "train_045698", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["evaluation", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "4de97e22bb56ba131756845c4a60bc517f23e3c3"} {"id": "train_045699", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 9,\n \"confidence\": 0.9,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["security", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "932d6f56f2753bb3e9f54b01b157a8e939e9fa15"} {"id": "train_045700", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["security", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "a09f5498f5ecdbe10d64ba7400eeb3ef2ca3bbd4"} {"id": "train_045701", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Python", "moe_experts": ["governance", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "6a5d3d3a68c30e9769c69ed721613e8fbb488a42"} {"id": "train_045702", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["performance", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "2f9fffe057baeea3859691bc1d1e0714b2f9c876"} {"id": "train_045703", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["evaluation", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "25608dd20f959381f925b3f2d2f6e49306b9c546"} {"id": "train_045704", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["security", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "a6f05e0178ffd311d280207ed0ce916097612714"} {"id": "train_045705", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "e8151462a12de65673dc5a15d132f4e7372f73e0"} {"id": "train_045706", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["security", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "fe28d994c76d494c8d9fd78df55513dba6ba7fd0"} {"id": "train_045707", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "e9062964a11095606bc5a2103edf7de9daa2ee1c"} {"id": "train_045708", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["governance", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "4c07dcf8c0b7bafd4bd6ee22daf437fb4d1bd44a"} {"id": "train_045709", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "5c3983c985d8507c380ab78f97ed982e712f7fc0"} {"id": "train_045710", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "c9b85f107a8e65dfdcd9c5eb452e455938182feb"} {"id": "train_045711", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["data", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "51376da82743e48d0383a0470ef81747a2886550"} {"id": "train_045712", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["performance", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "29e19804b05f616f785d07e17f13af20b468a384"} {"id": "train_045713", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "b596ff9b0af62fd2d8c5bb3744f97b672c553ee4"} {"id": "train_045714", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "39c8e7631adbc7794e409ce15e0c77fcfaa50a6f"} {"id": "train_045715", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["data", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "a914a3a3b8ea306b53f7a1bd9f6f8a56f4eeb6b0"} {"id": "train_045716", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["performance", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "b8bf08748500a1dd384f640bcd10f36a43bf8b93"} {"id": "train_045717", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 9,\n \"confidence\": 0.9,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["coding", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "04470fe2236a525c40de085326ff86dd35068ba5"} {"id": "train_045718", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "4b1475cace1af4abed8cf0cbe39501c3e2a04fbb"} {"id": "train_045719", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "9c5d944830f2517584e02815b961b6086525a4bd"} {"id": "train_045720", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["research", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "5d7b63df8ec7e4a3004ae6ba1c3792daa4594258"} {"id": "train_045721", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["evaluation", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "74051ad0e13869014310bef77457be593e9d26b4"} {"id": "train_045722", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["security", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "d9acc5892c6effa153ce310a876a0eb2dce75cc0"} {"id": "train_045723", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["performance", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "bcccc89ebf38e57919dafee549f787a9dd907e59"} {"id": "train_045724", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "c16db4dd542d681ed9fc055d74f8c8d2492424e5"} {"id": "train_045725", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "c5d0b6a72428fa21a4b004962e6507df6c5d34db"} {"id": "train_045726", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["data", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "95b0a8af2df7b2ef359cb45d49c70fd4719a677d"} {"id": "train_045727", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "a8a1db1dabfe472fc9e59a223768d258e7487b09"} {"id": "train_045728", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["security", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "626fdab58f2e27525f312243ff24f91c2eebf3bc"} {"id": "train_045729", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["governance", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "7660b1e5fa353feb3707471defde0ef156076c08"} {"id": "train_045730", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "e9f0a7461595fa44d4b99a0774eed7273881c75b"} {"id": "train_045731", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "8d20a7497aaa2edb4a4206ef083581f739f1d155"} {"id": "train_045732", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["security", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "8d73f03ec9d9531ffd8032ace198e6e5f2d22617"} {"id": "train_045733", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "54e67118d68733263b1585a9989f5ff380ef2f25"} {"id": "train_045734", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["data", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "e6cdec9a5418dceaf8645ebf67dadf8e99f34547"} {"id": "train_045735", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "C#", "moe_experts": ["performance", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "240462aaf8fe71b56565a10386e39eea42511e19"} {"id": "train_045736", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "bc233da0b8b6f4845e22a38275bbe41e5e217ee9"} {"id": "train_045737", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "0c4b164cf4996d49a7ea6b1d2fbde6e20ba18c4c"} {"id": "train_045738", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["security", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "9dbff9c3e6d94a7e747ef765f46ce7eb6f74f1f5"} {"id": "train_045739", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "e6cdec9a5418dceaf8645ebf67dadf8e99f34547"} {"id": "train_045740", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "ac2bbfdb44af3f03c1f7a6c46515f17249eb778e"} {"id": "train_045741", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "26bc4bf44da031211d1b19d9996400186f34f1e4"} {"id": "train_045742", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["data", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "4ca26d7b4d7f6c6b51a8225eed0d80ca3ee66335"} {"id": "train_045743", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["coding", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "0a9dcd82adf6f9897f5357e6abd70a18a9aae2f0"} {"id": "train_045744", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["security", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "18fd609639609c81d564d0b4e6cdad7b948f36ce"} {"id": "train_045745", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "c09e34ee3e56c21c4a66061200743b34e6cf3373"} {"id": "train_045746", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "9a5902f089ec1fe9723585faa0911fe39cf49ba1"} {"id": "train_045747", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["data", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "0d79316061e63ea30224bf85a2b7c51b67f18b95"} {"id": "train_045748", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "920ff6bacc7c0c419e21ef69131aa67ef62a435b"} {"id": "train_045749", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "6a8dacebc2a8959494fa2f9154b68b96f9451306"} {"id": "train_045750", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "974e65f54cb979a29065ccd6610caecce6d7238e"} {"id": "train_045751", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["agent", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "7689f86a3df1d28f7237955a68208eeffefc0840"} {"id": "train_045752", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["security", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "09a9e2aa56ea28ea258d224b2ee44a80f7700804"} {"id": "train_045753", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Python", "moe_experts": ["performance", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "2ca6310a8e70e1f721611a5354362e9add1e1748"} {"id": "train_045754", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["governance", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "38ba6f9455b9c3d77f901fca61c2eb39237e7b01"} {"id": "train_045755", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["evaluation", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "c4487ebd06b698a9846f94c7003a1fba5880066e"} {"id": "train_045756", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["security", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "6f250b8ff67dcfa4f2878cd276fe250089162021"} {"id": "train_045757", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "75f19f084855cdbe019301ad6dd73ce6e3f9a229"} {"id": "train_045758", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "81ac9a92638657745501e6ed7a8c89fd017b528b"} {"id": "train_045759", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["performance", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "f0f4e84aaafd157f0890ebf8062217d0d340b4b8"} {"id": "train_045760", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "c126946c4b0e0009e975b2834ea1afb1f97c1104"} {"id": "train_045761", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["governance", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "15e689c3d3d8672132739ded0e40130e1bb19d55"} {"id": "train_045762", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "442568b11bbe07813563f46d962eb0632669f2ed"} {"id": "train_045763", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "716bab34bf834a36bb2c8c2e69c23efe7f973195"} {"id": "train_045764", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "97cff8d8601d404e37d9ec81346be06b259fbc28"} {"id": "train_045765", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Python", "moe_experts": ["governance", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "f7392bfcb21a048e4d4b0bdbfeab7818cb6a736a"} {"id": "train_045766", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "e697a778d8612f70d8b6bbdf6a56704dc03db0ad"} {"id": "train_045767", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["evaluation", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "f6803572862187930deb3bc0b1b67c73abaeb9e8"} {"id": "train_045768", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["agent", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "a8c482ac82343cc789e033d88c1d3108e985df84"} {"id": "train_045769", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Python", "moe_experts": ["agent", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "d1d13a07ec3cb1dccc57be1026ffd44a98fff5f1"} {"id": "train_045770", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "5aebc4cddf0ef3e4ab831948eb22f2a9af8b6805"} {"id": "train_045771", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "b0791045ee1043ca69d9be0b45234081e453322a"} {"id": "train_045772", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["research", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "28df67f7f720317c99ebb8ab3a0c5bebdcc3609a"} {"id": "train_045773", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["agent", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "31e054cb2661f654fd7edcae423967043968bc8f"} {"id": "train_045774", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["data", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "f692890e518b886f54c56519cb72f0a2e637456c"} {"id": "train_045775", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["evaluation", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "b4886ddddbecf68d1edb118093dc2891986a9267"} {"id": "train_045776", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["evaluation", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "e08701f543b99c21e7db5c25fdd27b37ecee28ff"} {"id": "train_045777", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "d3a41ed8aebac04396166a4043a774a40001d7c9"} {"id": "train_045778", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["research", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "f2bbded922cd9dbdf0e1c1b29ac61dd7f672cac3"} {"id": "train_045779", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "e514092845226633efc4fe820f92c7e2f8eb2c53"} {"id": "train_045780", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["performance", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "3eb7c7dfee5e0f884c751abd56541c3bfaa0101f"} {"id": "train_045781", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["governance", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "f018518258480b0c96e94387d5850785af2ab9f4"} {"id": "train_045782", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "8a62e2b91b9ec02a1e5258c22a935b952d899458"} {"id": "train_045783", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "dbd3f8162f4ac5310355053e437c3eb5b5a2b82e"} {"id": "train_045784", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["security", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "eff785af56a4385e33fb37529c1a861a17f846fa"} {"id": "train_045785", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["coding", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "2226726b92c9e0677f91c6fd011802b762bf72df"} {"id": "train_045786", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "171579cfce59abe5786fbd7e983226ddf647a9c6"} {"id": "train_045787", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "0c643402afa1e78966b6a673a5c2e3c77a561142"} {"id": "train_045788", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["agent", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "b02e54b39fa8a7e4046b3d4ef435ef6f6ca02187"} {"id": "train_045789", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["coding", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "c9cd0fc4864dbb4e528bca17c6107b148c74b0d6"} {"id": "train_045790", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "fa8698f2e37c3a6145dc7834630bd8dc35a7ece9"} {"id": "train_045791", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["agent", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "d769c1e13198c3173c6e2c4d8e5d218382faaf9c"} {"id": "train_045792", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Python", "moe_experts": ["data", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "d23e3dfc7c65df4168e941a145a87e4afe9484f2"} {"id": "train_045793", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["agent", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "e6453511ff1d360cf2f30ee9aed61c457a00d04e"} {"id": "train_045794", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "43bd084e0927721632e8aecf85497ec19d900b13"} {"id": "train_045795", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "2752ffc5cc37938a6d9be3aac63b6e135121c756"} {"id": "train_045796", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["performance", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "33824aa12a42e85084efa85df31653d7013a8fe1"} {"id": "train_045797", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["evaluation", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "52fee6a0e8cf623e71ab7c11c80048835adec258"} {"id": "train_045798", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["performance", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "da187dc434fad34d99ea81122701cbaeabb2e175"} {"id": "train_045799", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["research", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "a64f2796365a21052c264ff225f031202cc5dab1"} {"id": "train_045800", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["evaluation", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "64bed467e608ac3700b8af699ac704edb93a6fa7"} {"id": "train_045801", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "5588fa6a5f99cbdbcfcf2b502cb2d73d8b641e9e"} {"id": "train_045802", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["security", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "6aa9787b717e436f91941652dd0250ad4e9ef3ce"} {"id": "train_045803", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "af683eca50879e67c6ac8be6ae6578fda552d566"} {"id": "train_045804", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "07bd75c93a6e9c11bf0c8cc2f7d863d0712d23d8"} {"id": "train_045805", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["coding", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "4b0c8584eb2192e98e3183c5a1b2884c38feca8e"} {"id": "train_045806", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "5c324614895768c8a670f48606864f76c01384f9"} {"id": "train_045807", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "58aaf0d8932d94daac16196fcba768ac88d60747"} {"id": "train_045808", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "020899d20a27a2fa5651ba40a50b9ac4bb6052ef"} {"id": "train_045809", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "47e942fae6f6453b15995754161fe5a07fbf3757"} {"id": "train_045810", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["security", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "43fed9166b8ffabfceb049547c7b0687da114fae"} {"id": "train_045811", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "b67c6d9a046c81ac956a829a2e72347599a229be"} {"id": "train_045812", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "c2427bf736afeb4f98eb1653b6c2f43a785ff2f9"} {"id": "train_045813", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["data", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "d9f3cea41eba3e92a9cc3f406391828e8ee57051"} {"id": "train_045814", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["security", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "888ed8e6b913f94a6e5a751dc7ea597d5a7df44d"} {"id": "train_045815", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["data", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "945994d2429904a0cea06fb02aa26a6a49ccbd70"} {"id": "train_045816", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["performance", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "fb391519d5591ff058c5c378a13e893658fa37c6"} {"id": "train_045817", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Java", "moe_experts": ["performance", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "ae75c0dd65fa08695e54ed88f609e0fa3e4b60e1"} {"id": "train_045818", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "399067b0189f378ef805b0364255a53c18e40b0d"} {"id": "train_045819", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "259e86ff7e200bc62c09b911628bb318f46a92a4"} {"id": "train_045820", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["agent", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "279788254cf318b8fd5749eb754dcdba01625326"} {"id": "train_045821", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["performance", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "35f7f3fbbe3065ea55fad45e40545d53cea0eead"} {"id": "train_045822", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["data", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "a12e7e9a34c0538f1ba47bb31419b1ef2ef6a2d1"} {"id": "train_045823", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "c868ad620e776f57dc0d9c6f4e9ee31ae79215c2"} {"id": "train_045824", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "b68aa8d6a99e9262d8dd62f20bace6f06a78bb96"} {"id": "train_045825", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["security", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "9930b2258527a1f9e171e008d0f7946bb08cfd33"} {"id": "train_045826", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["agent", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "63e8c951ca7f512054d911c3e02f66c2d3049a54"} {"id": "train_045827", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "4586e20b5bcee667d0f78854b361b66d1d4dde15"} {"id": "train_045828", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["agent", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "3f23d879a492914a5a3ec74e3b262b8be17fe652"} {"id": "train_045829", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Go", "moe_experts": ["security", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "75bc660e35f6fcaa46730fb1745dc6065b313ad8"} {"id": "train_045830", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["research", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "0dbfce0da27d9d2b39993a638153a86694644ab2"} {"id": "train_045831", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "bd2e8d705f09e9cd889039586ac6b4a005eba75b"} {"id": "train_045832", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "542d97acc90e4ab4f38c6633a5e91b34e7da8970"} {"id": "train_045833", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "16cdb8403a7ff233b5283ccfc70ac56b7cedb635"} {"id": "train_045834", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["performance", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "4b05aba03b84e8424a7852f0671bbfbb6cf8a67a"} {"id": "train_045835", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["research", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "4f128ace204a8041c6d993b44c42bbe39ec68c97"} {"id": "train_045836", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["data", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "3859be78d3c22c1bcd47500dfa72535db41b8060"} {"id": "train_045837", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "04d7c488eb0413115ca82cc3efc23380c70caaa8"} {"id": "train_045838", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["data", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "1850c90da7c75f5d0958bfecadde00e7474e52e0"} {"id": "train_045839", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "75df4fa54ac86caa3d576c3e0df50f3a3354d033"} {"id": "train_045840", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["data", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "9391685319931c0df4c28709360f77c048107180"} {"id": "train_045841", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Bash", "moe_experts": ["agent", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "075aa6ee201149592214955aafa9eb9e780489ee"} {"id": "train_045842", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "d5c8870c80f696e1bf399c4eb32ee63df7e403f8"} {"id": "train_045843", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["performance", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "a4905f742ddd8371362ac53b0ec33f549376290c"} {"id": "train_045844", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["data", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "a300be8952907cf69609cd8c38c4ab0675ba20a5"} {"id": "train_045845", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "0cbd10495ca9a2d6347014ac99654192d0e10f12"} {"id": "train_045846", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "bbc971daf3983d139ed50a1c168043ff5828d0c8"} {"id": "train_045847", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "80ce7b3b49e14ce43714682f97d921aed8c341e8"} {"id": "train_045848", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["governance", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "6355595532d31a3e827732aa4bf76019365601cf"} {"id": "train_045849", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["coding", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "ca46a3ae1a620460e94d42e8015eede3e93e26ee"} {"id": "train_045850", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["coding", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "e55896dc9aa4e86d51b0a092cc1df5ca0bf1af55"} {"id": "train_045851", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["research", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "a979930e9dcc827e30f52503ab51432c63f20a97"} {"id": "train_045852", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "97efb7fcda1a55c0000db9c8e1dae78dbd02d449"} {"id": "train_045853", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["data", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "5dd674849312b44618a894a2e38f6dce0cef93b1"} {"id": "train_045854", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["agent", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "fb645a70c140e21e5bb772aedaf8db715652ff20"} {"id": "train_045855", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["performance", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "86a07249eec76aa8519242ffd7df03c55d34e27b"} {"id": "train_045856", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "76904e222680e03266466135dd85aa06e2a742e5"} {"id": "train_045857", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "f6cb0a30188a07967894a0b6422d55895c5e4c6b"} {"id": "train_045858", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["data", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "58e6b6c0a2bdf6aa4c33689f28b72fa58641bba5"} {"id": "train_045859", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "c9bb2df5d2cf979c2de4d98aaacc1d67f76f1b61"} {"id": "train_045860", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "8b705a0d5e913bf9bea8dd670af4ae9567f7a1a6"} {"id": "train_045861", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "b86a227932de3a28ca5b1149cde1f524d5ff51e7"} {"id": "train_045862", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["governance", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "0ef3a293381b46d079005f84c1eec110908088a4"} {"id": "train_045863", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["data", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "626fdab58f2e27525f312243ff24f91c2eebf3bc"} {"id": "train_045864", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "2e34991ba666bd2edbb91351ae0831ba5e5921de"} {"id": "train_045865", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["agent", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "d713b201c0dcf6a6585e02d47d72db22ed805b8d"} {"id": "train_045866", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["data", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "dd30b5b5790ef88e5cb3795ba0fe44daf17ab656"} {"id": "train_045867", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["agent", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "557c59259d191b47662216f54eb8c9ea83455216"} {"id": "train_045868", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["data", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "eadb4e44dd0f48daa545d5fa0b570b7b79f1362d"} {"id": "train_045869", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["data", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "5f43a2d864b3109a5bd7fee88596227d201478e9"} {"id": "train_045870", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["data", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "159d9c89bb08bf18de9038654e3df166b6e7c05a"} {"id": "train_045871", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Rust", "moe_experts": ["performance", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "692c4f0b6b506603eba711416d5f20c9b276c164"} {"id": "train_045872", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["research", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "97efb7fcda1a55c0000db9c8e1dae78dbd02d449"} {"id": "train_045873", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "a84e80be3d6941c6fa3d86a61496490f6ad45e0a"} {"id": "train_045874", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "588294c820470349cc13e91be0b756f5bf782093"} {"id": "train_045875", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "b61425a320a3a8ae096f733d2e8b74e285058153"} {"id": "train_045876", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["governance", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "a8e84969c2bbf7a1bf11c05d3ed38fc570e8b614"} {"id": "train_045877", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "827ac587ec6d47e1d4c56de18efe270c11b34290"} {"id": "train_045878", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["performance", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "969d832e43a71ec58e885144ff560db45ee5fafb"} {"id": "train_045879", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "268bbcff3fc7e7d43417528030467529d89a5832"} {"id": "train_045880", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "7d5287bbed42a8370911237a00f2a3b1c70ef68a"} {"id": "train_045881", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["data", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "f0a72273b93aabd026fd101741060f2f1ca1fb06"} {"id": "train_045882", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "231344d7d4e3a0c7aed09b436137715b135bcf83"} {"id": "train_045883", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "f2cd6b3d65da1eaaf8c6446e072a20cdc3bfb7e4"} {"id": "train_045884", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["research", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "72bcb681f614f2f27998de4881972dedd96614a3"} {"id": "train_045885", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["performance", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "539efebcf19c424eeb67a9fc614083ab2f30c146"} {"id": "train_045886", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["coding", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "efcd78c1ac745a88be55613d1e032358929c48b8"} {"id": "train_045887", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["governance", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "9552de665f446d34b9e93272cfad5b007fd8e28b"} {"id": "train_045888", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "0904302c334c8a49742814df697f3de0d6e09764"} {"id": "train_045889", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["coding", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "2f4ceebd2697b921b04d334d57ada19e9de229f6"} {"id": "train_045890", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "459ae868a93d8126ff71e746923764048d82c38a"} {"id": "train_045891", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "a9edfbeca8ee9dd9e87bb4c7872dd8182409244e"} {"id": "train_045892", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "5d2d3bd14e5e24f89e46a0763f302010ec96ace6"} {"id": "train_045893", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["coding", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "1c5ccb667bb418bfe2364da5c4649bee6fb67797"} {"id": "train_045894", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n cost = 0\n for _ in range(budget):\n p = plan(issue)\n patch = edit(issue, p)\n ok, _ = test(patch)\n cost += 1\n if ok:\n return patch, cost\n raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["governance", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "f5d5fca7d9335e2867172619b4a07969ab3e7954"} {"id": "train_045895", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Python", "moe_experts": ["security", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "9d857058e03df2e6da3b262f44177f08c3e155b2"} {"id": "train_045896", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "cb69aa3c4ba85e61726741b98b3ea0689ef3cc6f"} {"id": "train_045897", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "6f02359e139ef17e66177b22350b8332c27c2396"} {"id": "train_045898", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "72426320e690740ba1eb40f49af7fa0573003732"} {"id": "train_045899", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["governance", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "7435169c58e19e03fa06ba7f87f7ffb18ace531e"} {"id": "train_045900", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "8d2329cbab7a2df71e19b6f69fb898a0dfd8fd80"} {"id": "train_045901", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 9,\n \"confidence\": 0.9,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["performance", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "924f4373aced7e0cbc68d1a2c026e0c5e36f39da"} {"id": "train_045902", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["data", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "cd24dbf1f824742256112411e9459bc894d02614"} {"id": "train_045903", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["security", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "bfd9d51c689651a4d914a8fa6a224ea1e9be83ff"} {"id": "train_045904", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["research", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "78d8ef7bedee93e9f570ef80312d46bad7c8c636"} {"id": "train_045905", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["coding", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "eb7cd50d0fbb96edef67028a88eeb6d3cbddf697"} {"id": "train_045906", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Java", "moe_experts": ["governance", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "42a93429fcd1fb8f5b1bb4b91a329c228c117ec2"} {"id": "train_045907", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "2ebac9e136a595748fc8c8852a941f0f65c694f8"} {"id": "train_045908", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Bash", "moe_experts": ["governance", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "a31cf1ac69ffb5d5917c3514ca2538c41bce2ebd"} {"id": "train_045909", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "063468880cd23cf11bbe41cfdb3d836311d13be3"} {"id": "train_045910", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["security", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "75039968b44b8e5ba9d3f5f019c4ccd4be578eb8"} {"id": "train_045911", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["security", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "00642de36870c5efb087a41fbf911205e347db64"} {"id": "train_045912", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "348fe4e0e893334d269e3e4c9adf647b24814766"} {"id": "train_045913", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["security", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "b0816926d3a12abb764f63e4134e41ea32f170d3"} {"id": "train_045914", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "6b1be8834843489802aa8b496b37f3d9442bf7b4"} {"id": "train_045915", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["governance", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "31cddf44315823cd388d421211f37df346880e1a"} {"id": "train_045916", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["security", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "9e75b87ee55cf0c16d35cfebad08a3e3d1dc4a7f"} {"id": "train_045917", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "c16fe1ded1c65e55aa0ba0e26d09b6f143ebcc3d"} {"id": "train_045918", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["performance", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "c31abb935b2cdcba5dff0288b5b398b8c347e716"} {"id": "train_045919", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["security", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "58bc280d3a8d8182f86aaadc40c13e593a7df6a9"} {"id": "train_045920", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["coding", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "54c809d44001c964aa65bac51393fd330a4dc43a"} {"id": "train_045921", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "33872e5bde25fbdd2e3273765cf227484aad9e89"} {"id": "train_045922", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["data", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "33d31a21ae9ef7b34f84ef112139adb4fd530061"} {"id": "train_045923", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["coding", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "42168fc60098e4157b3433646f192521d706e916"} {"id": "train_045924", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "773d4f13af0deb90de69dd8ac2cbf49a0eb0e920"} {"id": "train_045925", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "59aa170e106b077416b118b92a9d59a54314522e"} {"id": "train_045926", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["research", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "bfebad7db4d8db4d2d667dd02d27f657c7a57a79"} {"id": "train_045927", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "8609760df0d4e3215e60e53e835e615d161c2b56"} {"id": "train_045928", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "fb5eceaeeb86ce3df4c82c712251052582f19e3a"} {"id": "train_045929", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "a4e1b2f23447622ad79229f1f68e764a0ebdd93a"} {"id": "train_045930", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "fa8a6b5b0703f58942633dcce35e1c0622e0f507"} {"id": "train_045931", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["governance", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "a0a4588d1b30617f882e11011362493bb5212c5a"} {"id": "train_045932", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["agent", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "13268ad56644096213cadaaf9073568a02fbef4d"} {"id": "train_045933", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "d9c4e35b1392fd0347e3f1f3cf8bbc9d1183f3b0"} {"id": "train_045934", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["evaluation", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "3d408bb18f315ce30725f226f3103e3e7addd271"} {"id": "train_045935", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["data", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "7d036c28559e6091b60edff3a61c2156014572c9"} {"id": "train_045936", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "aaec72989924dd3cee917f263d021909273cce20"} {"id": "train_045937", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["security", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "c2102c2aeea58f42f30346234f790eb9679ffeed"} {"id": "train_045938", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["evaluation", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "0821b4565a8050b5a6efe5a856ee08196c43d7f9"} {"id": "train_045939", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Python", "moe_experts": ["performance", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "fb37479fe07da3d3a37da55892adefc122b8e8d5"} {"id": "train_045940", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "45f421a2c0f4e4cb8f149c159c3d6236c2bd19a0"} {"id": "train_045941", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["research", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "1a18e489e761f82ad2a562dc25d99fae80cfa961"} {"id": "train_045942", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "7d66c54f3b86f8dbe096f5c737b6f89ba6eff777"} {"id": "train_045943", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Java", "moe_experts": ["coding", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "d426c8a9d9f78c8596620e73ff70596ab1e39c66"} {"id": "train_045944", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "da5d39d9d8bc58f0a1d198fe39d18c2ad7b75732"} {"id": "train_045945", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["performance", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "6317d753841faa842b3f82fe907a17c6eb48da02"} {"id": "train_045946", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "aaa059099ca34f470740dc63d5cc35a2faf30a14"} {"id": "train_045947", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["agent", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "5694651f444b66fd8f1dc2592d7f5940e58bdd6f"} {"id": "train_045948", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["performance", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "3d7320aef923c10f9d26af8b40e5b500728d1ff9"} {"id": "train_045949", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["evaluation", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "86db11b10ed2dc4a5710c776381bd19d6fe4516e"} {"id": "train_045950", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["performance", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "4f0d1b7ef9f6ab81dc390b7b88366d8d3367a269"} {"id": "train_045951", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "a11cdfb44cafe8d50bc471f464e5ff96c13da2f0"} {"id": "train_045952", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["data", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "25d90f6ba63dabf6eb44d55e836ef6281e09f1c4"} {"id": "train_045953", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "4aa7794709546ff4dda528109e94b9249aca49d6"} {"id": "train_045954", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "16a8323b977f54768bdb91e162a72edc8d50bf2b"} {"id": "train_045955", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["security", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "9c959812936dd0b622aa47821a2fb86559453071"} {"id": "train_045956", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["agent", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "22e5bbf4414256f6b4ea6e838356809b5a936810"} {"id": "train_045957", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["coding", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "a70aa8a69e0cb1acc055f9c5da32748bb81ae23f"} {"id": "train_045958", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["data", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "cb7628838ad56508eabb38bb56cf23cbac19d9ce"} {"id": "train_045959", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["security", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "c669a81f11cd84ebdc5a680cacd9c6a9d6739a48"} {"id": "train_045960", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Python", "moe_experts": ["data", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "39349d4a8c5fb334fe0ca9f56dfe7b9ab8e682a5"} {"id": "train_045961", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["performance", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "34f7817732d33526ac86fef3eed3e7d4dffc3566"} {"id": "train_045962", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "4b209c59cef8a3fcfc5097c9ba887d88fdc95073"} {"id": "train_045963", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "20e09f2d4b0c4664043dad12951500eaeec8198c"} {"id": "train_045964", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["data", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "7d036c28559e6091b60edff3a61c2156014572c9"} {"id": "train_045965", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "4808195238b02741a93c44dd77cb410c2a3169af"} {"id": "train_045966", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["performance", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "717d1f54a9e364ff5a30eeeb0226f7ea3b7c1c50"} {"id": "train_045967", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["performance", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "40f34e24cedde2819a3d02828bbe40b2eee387c9"} {"id": "train_045968", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "874fd218fd3d5dd9b95bd30bb3ac6922f2241393"} {"id": "train_045969", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "a5170a6f03b47e09d8bd2ad9fabdd64423d4d302"} {"id": "train_045970", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "fa187b39eb20deb39785fe15808cc50a4d38efec"} {"id": "train_045971", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["governance", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "122aeff34979aaca88578640582983c02c4705e0"} {"id": "train_045972", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["governance", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "e1ca443eb1d5f59e5256cfc173b8ad7e41c351ba"} {"id": "train_045973", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "0dbd99f2978d7fb9caa103c6beaf69d422d2c574"} {"id": "train_045974", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "9f1778a91431c3da438d5c438a6889755c696947"} {"id": "train_045975", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["agent", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "6370f99f6b4d3c8ef3892cd5d4f1ac00b72d281b"} {"id": "train_045976", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["security", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "9747c08afbbbb92819b01eeab48f10439894a656"} {"id": "train_045977", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Go", "moe_experts": ["security", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "7adb10256b75668b071b5b6fa554991b1a6ebe89"} {"id": "train_045978", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "d7fc1f80b1986aaa01242bac36f9183df8006d3c"} {"id": "train_045979", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "3839c779f43ec5527a05bffa7113bfb2105f82d0"} {"id": "train_045980", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "23a0b07a29204c9edb32ba077dc239f355ad8243"} {"id": "train_045981", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "74ec87ec2899d6ea4842aeb57380e728b8218566"} {"id": "train_045982", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["data", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "ae830ea1e6dbdf806cb7c164581a2bd0cff646b8"} {"id": "train_045983", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["coding", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "701a036a00c5d6f0ca60ccd2f1504e0d8e49c9e2"} {"id": "train_045984", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "76d8892282bccd04220b3527524748bff67758dc"} {"id": "train_045985", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "65d996ce87d918309595e6aeca3ce88a3750ca98"} {"id": "train_045986", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "2ab17f50df5ee830d501d849dd9d3f6f43c3a24d"} {"id": "train_045987", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "63892054bb80f18979ff1db2f0391538fcd8fe79"} {"id": "train_045988", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["agent", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "d02d4c88b14397b3e0a6e4294a0dbd7bab07dbb8"} {"id": "train_045989", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "574b97cd5fd7f5d8f37de58174c27204378324b5"} {"id": "train_045990", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "71b1a955b1d9c74128c4e8575e4aad7c154575b2"} {"id": "train_045991", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "430116401fb346277a97f1daa4a45cf820a0466c"} {"id": "train_045992", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["research", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "4b7ec8537200358fb69577fb16ea23b108649a67"} {"id": "train_045993", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "a5d6a077ae8b7df04ccd68a476cccba728adcc38"} {"id": "train_045994", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "b2367d803c6fa52506b5eaac7a56add036bb858b"} {"id": "train_045995", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["data", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "7b25e8f5a65c7fa5d457e8a6809dddd9d96a6355"} {"id": "train_045996", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "SQL", "moe_experts": ["research", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "8fb0a33b3879fcc4d65c562a5f7a8890e85e7092"} {"id": "train_045997", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "3d53ab34b98a1e6ab354a81d54fb121b2cdcc978"} {"id": "train_045998", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "278923d979910313d5087c5471c392c3bafe7a00"} {"id": "train_045999", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["evaluation", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "f10afed923bcfe9665240e0a41f3d09d813f902e"} {"id": "train_046000", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "28652ad509709a726ad26ef8850e194ae67b8f67"} {"id": "train_046001", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["research", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "5ec2e0080d780c03dd6ba1b1e4321d83a6c6b448"} {"id": "train_046002", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Java", "moe_experts": ["research", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "7822553c4e576513d4f5940cb50500f42e84ad6d"} {"id": "train_046003", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["governance", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "fcc25575bdbe6f2cab48d82547764ae0245260ef"} {"id": "train_046004", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["data", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "fff382a94cd2f1eef484e468ff94f5f3a0e4c2ff"} {"id": "train_046005", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "24afd6123066a29154d56c7608cf5a58ea9ddae0"} {"id": "train_046006", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["research", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "99fdd164420e5eb2c4cef9b44ada8181ccbc6a3f"} {"id": "train_046007", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "a099657211ff03839238f78ba0d8928f40829010"} {"id": "train_046008", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "47dc9abfb3f02aee0787d4d00047f812f70f55fa"} {"id": "train_046009", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["agent", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "8214a4354fb73e0efaac5531bf80b8e037a94135"} {"id": "train_046010", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["security", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "464a944e80d529187e1f84ef4d0298669aaa4891"} {"id": "train_046011", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["security", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "c35de672765fc518928c9af322ac39e6e6a9ee37"} {"id": "train_046012", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "bcac6d0fbee6cf2a0b4fa1b7579788ccd4784bbd"} {"id": "train_046013", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["coding", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "a83fc5c9139876b0b6214ee215d2421e8a75bc0f"} {"id": "train_046014", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["data", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "c32bfa5002cb0f3f5c2b64f67099e93e23ff1274"} {"id": "train_046015", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["data", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "354fafb4890c3cb2a5a413fe5c9c9ecbe8b32be9"} {"id": "train_046016", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "77aaf47b35fff9adf3321ed8625bb8b7c42428ab"} {"id": "train_046017", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "528c62a8d8d16eb6e34f435a2c1274882fb41187"} {"id": "train_046018", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["data", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "f9bc3203723a0bcf476051ecb955e9591ee18579"} {"id": "train_046019", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["data", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "92a69375df9aa8c64348537491a576dcd36a36dd"} {"id": "train_046020", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 9,\n \"confidence\": 0.9,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["evaluation", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "53b5d9115610b360c4411b4ca892b1032c458b66"} {"id": "train_046021", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["coding", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "9fcde85cf9d38c447328b68e5987edfccb36437d"} {"id": "train_046022", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "ae2c42f0ce7dfabb1412177c59324c60713ff580"} {"id": "train_046023", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Java", "moe_experts": ["security", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "90abe3c34e80ff67d2dd27be164e5b1fb29dc12b"} {"id": "train_046024", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "55aa955566eee6ef96347037aa598c63e99b6b08"} {"id": "train_046025", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["research", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "a282f539fe94307934e1d646b81b78e6afa17e5e"} {"id": "train_046026", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["performance", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "7ff2fbd7c9ddc42c68908a23d53d803be8ea3979"} {"id": "train_046027", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "72e7ca0b393fcf81d7da4ea70d56c99a75459fa9"} {"id": "train_046028", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "5d4315dc9f90711b9b608b4d5c09db81f38f491d"} {"id": "train_046029", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["governance", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "802c77246ffcbd44c2a6fcba42e945888f589158"} {"id": "train_046030", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["evaluation", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "fe886960abfec00f4dd64774d73a378787731da5"} {"id": "train_046031", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "9eef229fd9a18a1bfa611e272743a82e41fd7a5f"} {"id": "train_046032", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Python", "moe_experts": ["coding", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "cde6351c43fa713b69cd7b64e9d4dba25661f815"} {"id": "train_046033", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["coding", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "3197ac35e5831950184b5bce9f9e0f5091340fb9"} {"id": "train_046034", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["evaluation", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "dd50c5245af94ba7a2d6e0ea62ae58137d1bc12a"} {"id": "train_046035", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["research", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "de01425e5288a52a1f298c434372cd3faca25692"} {"id": "train_046036", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["security", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "093f0a478260c6565167948f26ae503912764c7c"} {"id": "train_046037", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "741196d184af649daf132e5fac753784a84b30b5"} {"id": "train_046038", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "ae2f83e0a453f3ea93710a5cc384257cbc61cff6"} {"id": "train_046039", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "56a90080c1cdd99a93fd6ee402156a23db4fc947"} {"id": "train_046040", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["evaluation", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "7c6deb07e586df89756fa9938c032cf7f3e5a7be"} {"id": "train_046041", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "3298760228bbe4f94eabb0a212b34992166bfafa"} {"id": "train_046042", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "6b1018b4be3ff30d1af5b70dd896a2e87d021b94"} {"id": "train_046043", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["data", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "db55fc88461b4a5ece608429b8be200d179fdf38"} {"id": "train_046044", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["governance", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "7222c2e2386456404e4cdfd85cb57b81186b5037"} {"id": "train_046045", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["performance", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "ca980fc195d0ea84e1bed2aee74eb5c597f7f69a"} {"id": "train_046046", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "6b58325991dd3f7163cbf5895582d14dbb545322"} {"id": "train_046047", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "218be2544a84c6489cdd34e65a767c9fe08519a5"} {"id": "train_046048", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "26df428de712ac4a719a0dc304cc2bf6866489bb"} {"id": "train_046049", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "SQL", "moe_experts": ["security", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "cf3f565fb1bb923583eaab5cdd11aafdacf746ff"} {"id": "train_046050", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["agent", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "508aed416fc204530bcd96975dadaff7b459e652"} {"id": "train_046051", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["data", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "8c4c8e7000d98e38ae32883ab5b24c4fbc89f689"} {"id": "train_046052", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["evaluation", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "9b2c32b8e22715e5a87ed2b12d03a56518eeb890"} {"id": "train_046053", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["research", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "5347f1fe2783e5cdb98cc821094307d24c7a082f"} {"id": "train_046054", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "9c23957daea291a5351da27c2e7a1e66efc81bd5"} {"id": "train_046055", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Go", "moe_experts": ["data", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "f34b60000f1b2ac78e733b51d7b0ae5d27b4bd2f"} {"id": "train_046056", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["governance", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "15f42d1872a2bce757774ff2b0b21688c006ce9f"} {"id": "train_046057", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["security", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "0e0ea7c7037a9b572fab5951cbed0bea77ac9b52"} {"id": "train_046058", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["coding", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "1042a591717d25e76bbfc6c94d069a6c82751a65"} {"id": "train_046059", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "fb24dbdcb44a0e24f2abc7e3089a2fa585096ff5"} {"id": "train_046060", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["coding", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "c03e1fb6ab728f4000acee2432697ea937cbcec8"} {"id": "train_046061", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "95a4e85a405d61df531bbc34b9126059b33f8244"} {"id": "train_046062", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["performance", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "0edf02213cb2f8e6382ef7ca720ee722a094c968"} {"id": "train_046063", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["coding", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "699bc3cc4590652acec49bc3bf11872b200c3ec3"} {"id": "train_046064", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["evaluation", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "c69d709438c534ce9dabec030f6031e2f37c519f"} {"id": "train_046065", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "2ab0f561de78abe4da10f6793436b609807aa451"} {"id": "train_046066", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Rust", "moe_experts": ["agent", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "bf0a15660495dba8b004d6a3d9e1c0bf49b31699"} {"id": "train_046067", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "da6371b5c5440d96a5a2c88d454f406685902d94"} {"id": "train_046068", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "3f7df3140f55ec53fc7f2e59e767e60e9e6b70f4"} {"id": "train_046069", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Go", "moe_experts": ["agent", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "5d02d4731008f93e996a3b97acb4ee4b1195675e"} {"id": "train_046070", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "3c103bf558769fbdd679def2f2280458134240f8"} {"id": "train_046071", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["coding", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "059a5ce9e7f4eccb31f99df1636a90b87c004961"} {"id": "train_046072", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "6320ed841e49acb45d7c99bbdb031be2925d7520"} {"id": "train_046073", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "7aeaf9e9fafb5b1d883926d80b060e410447a6f7"} {"id": "train_046074", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "C#", "moe_experts": ["agent", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "51aaad3eac4d8cc62fb0873dafb426deff885b90"} {"id": "train_046075", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["coding", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "f0e223698f55a3d9c8d29a0c6d98d6ad52371f0a"} {"id": "train_046076", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["data", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "77a4c791b629cb23ee06a6227b23a1bb533bcd7f"} {"id": "train_046077", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["research", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "a0ea08be45e61915669f9865d5799986cf9d829a"} {"id": "train_046078", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["data", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "5617146f0e507f6e8b90cac54158b6f193e222be"} {"id": "train_046079", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["governance", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "059a5ce9e7f4eccb31f99df1636a90b87c004961"} {"id": "train_046080", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "273c75c945be9cae72cf17796b40524f4d8ac638"} {"id": "train_046081", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["governance", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "bdd3cc959661d05e46029b6494692935af5765a2"} {"id": "train_046082", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "c932fd7a65f0f0fb2aa81d2d434dd5469a3fa8ff"} {"id": "train_046083", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "C#", "moe_experts": ["performance", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "8a891f4aca0f1ffe0f7bf438120b727c2956e754"} {"id": "train_046084", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["data", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "67b7596874dd8e2a15d6281fb4c43f3f604b60df"} {"id": "train_046085", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["security", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "cd48e369517186753fc02525e2449866abc628a9"} {"id": "train_046086", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "bf9341ebe4ea2bae14358bb4b3a91999b8343369"} {"id": "train_046087", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 9,\n \"confidence\": 0.9,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "0777c9eb03db93f292b5307f9b9370d11c0451ec"} {"id": "train_046088", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["data", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "74d222ec036227166898c4b87e7e630c990fec73"} {"id": "train_046089", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "2532b57060da07ca5cffd3b189155d82dcd64d83"} {"id": "train_046090", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Java", "moe_experts": ["coding", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "9052e8cbee24c1a72be027fa96a57c3c7bf80049"} {"id": "train_046091", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "a290f24da17447d13890140e7aab503709334fde"} {"id": "train_046092", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "40da9a612831e0b4605ed29a6012497ebee38678"} {"id": "train_046093", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Java", "moe_experts": ["governance", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "665bf1683979a8bec8ea2cfec724148f973a8e81"} {"id": "train_046094", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "8f8745f06bd82353060f7eda63738ee194a2d85c"} {"id": "train_046095", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["data", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "d296e7e77079873d186204de0af718ce32131b70"} {"id": "train_046096", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "39ed88344489a4e822ac85da450f46bac25e44f2"} {"id": "train_046097", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["agent", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "704c9d0f524080900706c70441861c65a7fbf2a6"} {"id": "train_046098", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "a8eca372c658a4444b7277a758b4c7f92b70edf3"} {"id": "train_046099", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["agent", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "0d7d9386b2175b90174ecfd435b0f5492d0cea1e"} {"id": "train_046100", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "2ce9011a3b85b4e9e65d165d45def055651899da"} {"id": "train_046101", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n cost = 0\n for _ in range(budget):\n p = plan(issue)\n patch = edit(issue, p)\n ok, _ = test(patch)\n cost += 1\n if ok:\n return patch, cost\n raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["evaluation", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "24cd7ad76a0621b6a5f91e8446e12267593d04db"} {"id": "train_046102", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["governance", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "3fdb26dd14197d6a93d7e80a11880377c6475ddf"} {"id": "train_046103", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["evaluation", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "709110176498bfb261bb3ea411390c420b856f02"} {"id": "train_046104", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "ac6453bd6eac93869cdcf7b4d09cbe8082c168c1"} {"id": "train_046105", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "SQL", "moe_experts": ["governance", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "96628281e5333dec2e9c5d6d10406ee4ef4c2fc2"} {"id": "train_046106", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["agent", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "182a7834bead98b6bb51a41749e202e7f992d082"} {"id": "train_046107", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "3c2f9f7a8713bdf6a6c303a0c664e4714f70e0c8"} {"id": "train_046108", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["governance", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "45c8d503212f49b9bfcd75b0fa2a6defdaf6f1bc"} {"id": "train_046109", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["research", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "4cd01657d0e6a03dc6b2447189518adac568dc1a"} {"id": "train_046110", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Rust", "moe_experts": ["performance", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "c9d1f2761f6f9a39c77dcedb534a3c39e0ad31b6"} {"id": "train_046111", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Go", "moe_experts": ["governance", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "3b03b57964c8a5b512dc5cee7e0fbacc298c29ef"} {"id": "train_046112", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["governance", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "dfd07d1777e5104d356e6f98449921a4bdde9596"} {"id": "train_046113", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "7476416111778e2e46e4627b0019a7264ac6c528"} {"id": "train_046114", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Go", "moe_experts": ["evaluation", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "790c643c70a0de7682bdafb4b01f3a8222771d79"} {"id": "train_046115", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Rust", "moe_experts": ["security", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "f7c1cc0891f2abeffb633ed6ab7cc43df9025bc1"} {"id": "train_046116", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["security", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "53ede357c46e655b971be71ca83cde3dbfe71d3c"} {"id": "train_046117", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "8210f02dc9813ccab5e225d4e99811e735801be2"} {"id": "train_046118", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "0bb10d23c22daa7b5b300ff34c94c8b583187344"} {"id": "train_046119", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["coding", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "0dbd99f2978d7fb9caa103c6beaf69d422d2c574"} {"id": "train_046120", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["agent", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "0964276e70e255b3712878dff8be681fcc1f70e6"} {"id": "train_046121", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "a6b1b2a33d0b4faa41fc4f9f203ed8128b9eaef9"} {"id": "train_046122", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "40fa8657846abce04bd5b2bc9a09442c328b2792"} {"id": "train_046123", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "448e183ca04d159235d9ad4ef4ab72c26d0e36e0"} {"id": "train_046124", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "95987161142496ec33c9a304c5fe86fdf83781d2"} {"id": "train_046125", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["security", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "ee76ee93610386afa7859399fc657a0bb41508b1"} {"id": "train_046126", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "6b9f5f817311b00d72215e2c6dcbc81d57cd8d2e"} {"id": "train_046127", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["research", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "047638ab34e739fa9c22d22386d222daa5a0db1b"} {"id": "train_046128", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "6ac49a89a411ba72bf9185bb6c055de9ef228354"} {"id": "train_046129", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "ab945bb8735aceaf1f4955391676482367cef1ec"} {"id": "train_046130", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "916f8895aa83c5ebcfed036411ddc8af76ccfe20"} {"id": "train_046131", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "07a0ef700ae901af5bced86a0e344d0fc1abc4ba"} {"id": "train_046132", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "2747f4051adc81b55d7adbfddd4f03864409e192"} {"id": "train_046133", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Java", "moe_experts": ["data", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "90abe3c34e80ff67d2dd27be164e5b1fb29dc12b"} {"id": "train_046134", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Go", "moe_experts": ["agent", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "658921bb748cde791877817aa951fe6e9f18a987"} {"id": "train_046135", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["agent", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "33c238c4be49a3e301c5a7d8e2fb282c714b696f"} {"id": "train_046136", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["coding", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "2179ac5e01f8c77d4921c595386837e2b4ab5168"} {"id": "train_046137", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "efcd78c1ac745a88be55613d1e032358929c48b8"} {"id": "train_046138", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "9348c91e8f1bf4dfd6f3a151a49ede1ff536050b"} {"id": "train_046139", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "0fa41e369f5ce61f0802a8deaf44fca2f1928c66"} {"id": "train_046140", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "1aa832dd82a79935d9f6c034bbd6ba235e690d44"} {"id": "train_046141", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "f1b160bd15f08ffa3328092e5776f31afba73d12"} {"id": "train_046142", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["research", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "d1ca85545d1749004cf5bf96c5660945922c27f0"} {"id": "train_046143", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["data", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "9a5a6aec6f94d868b08aafe62df2f0217eeac9db"} {"id": "train_046144", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["performance", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "5c8d884c5ee78c844a86f9a20d4ed72e9a91dbb6"} {"id": "train_046145", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "dd13de73b164d2422542540f669751027a17c3fc"} {"id": "train_046146", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "768fa6bcfb11043314a78d8a10c75ae5e4ad9eed"} {"id": "train_046147", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "b0c3f48e81c7dda2b99d95f23a989bcb97130d78"} {"id": "train_046148", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "ada64629b4d38780bd1088baeeafcd48efaeab27"} {"id": "train_046149", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "47aaec5058a77ee58f48b45ed51d4600956aa3d9"} {"id": "train_046150", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "2d92c400bf3dbc0e15f404642de874807aae732a"} {"id": "train_046151", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "cadb0979d1b540913665b85222a172d82b8e3822"} {"id": "train_046152", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Bash", "moe_experts": ["research", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "d6cd6059c1b00e346665ea366dd85629cb09904b"} {"id": "train_046153", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "a0840c9572150c70a1c33a57c8f0e5aaa1d0c237"} {"id": "train_046154", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "b3d46694475e9485d77428e05e75b2b56716e5c1"} {"id": "train_046155", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "bdd0b471fa22e1624ff845186e91f4728e7dd9dd"} {"id": "train_046156", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["security", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "480127f520b23062eb230c69722e53b8ede4b219"} {"id": "train_046157", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "738bbb0934c65596503a183cc0ceb9a9c3b48583"} {"id": "train_046158", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["security", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "262e3f6295948cc261fba1b395dba20deb4263ba"} {"id": "train_046159", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "2c52efcb8d0be30e5499aa35762448f5a0c9b3f8"} {"id": "train_046160", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "25f99390f3d84fca5fba73f52996c505f92cd922"} {"id": "train_046161", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["data", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "9c439c236900f2252b11c187e2b1863a2cf760cd"} {"id": "train_046162", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "049bc3ad2fe5bed8f10acc973603dda6f57b80f6"} {"id": "train_046163", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "0c1f06af552e67875c73e60bfd4d7fdf6ca0be4a"} {"id": "train_046164", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "1d6a0fd4651dfddd7ee23ed50ccbed6ae4986a7e"} {"id": "train_046165", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "d0e4f7885824c5f666b8e2d52753ca12111d8368"} {"id": "train_046166", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["research", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "273261a4ce4511d40a7d0ca3879bb1d9d1eb82a4"} {"id": "train_046167", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["governance", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "493888bb07e89a38affd775d519a9d0c67e82bc8"} {"id": "train_046168", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "fc9b13f213c63cf6559ecba9e44affd4f46bba14"} {"id": "train_046169", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["data", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "e27ca4bcb2215169a776a3f4d366100128705b85"} {"id": "train_046170", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["performance", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "6c47747346fb6c55f87245ff9ceba844f43e4aeb"} {"id": "train_046171", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "bb2377646bba88835a67225bb50df44d7d3dac7d"} {"id": "train_046172", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "b3c7564f745b9cd69cbb72c290aee29c8675f996"} {"id": "train_046173", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["evaluation", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "c9e8081ccd358ec331de8ef68d4bcaf13e74c695"} {"id": "train_046174", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["security", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "7bfa268e318e5eb532d6d1f11cbacc23b79a3f7c"} {"id": "train_046175", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["governance", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "ab211ca2faab6ce2faf6642efb3aa4a861d07d8d"} {"id": "train_046176", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "SQL", "moe_experts": ["data", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "2c2482f841ed68a5c4d6c8cb3248dee2eb17e5df"} {"id": "train_046177", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["governance", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "9463465f3cc5cb478702f457a821e664f32f7876"} {"id": "train_046178", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "b5b97af04c6b2d33bfd95217f16f50e8adb5be86"} {"id": "train_046179", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "745b87d1d2c8dac9da5873e40668ae4b78cfa51d"} {"id": "train_046180", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["data", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "50ef229b0bbdad1f0f760eec1981b1866aec73cc"} {"id": "train_046181", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "cb073957286410ee6e934895c28f031cfafe2b83"} {"id": "train_046182", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Go", "moe_experts": ["evaluation", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "e578865f54a1ab8857234e8733a3e0190ade8f40"} {"id": "train_046183", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "30c83437867167b11341a482570ea48656e770d1"} {"id": "train_046184", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["governance", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "276d9e140a197874bc8562b5cc38616234427b85"} {"id": "train_046185", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "2d3bc534426984756cb2877672d38d255773d929"} {"id": "train_046186", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["coding", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "28dac7fc841579c745c98c6c5629402765b37a1f"} {"id": "train_046187", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["coding", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "8e44a822d6e30b0c81df18fcb68e3bc9e9eae1ac"} {"id": "train_046188", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "ac89b826a0e9c1cc61eec4ce2d6699ddfc114a8c"} {"id": "train_046189", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "40b62c11f228fb3e70957bf25848f22daf579cc8"} {"id": "train_046190", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "c273db0ee1d9763ff3c9bef6e1b2fc7cdb6754ed"} {"id": "train_046191", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "a4b7077f05f22d901a97f9bd9fcf4a583a991dfa"} {"id": "train_046192", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "bf9341ebe4ea2bae14358bb4b3a91999b8343369"} {"id": "train_046193", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "833d33c097cd8656ef60bff6633d794c808f165a"} {"id": "train_046194", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "9004dd994be2c04e08d128fc1d7233eeccd775ba"} {"id": "train_046195", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "33f55b1e963c0e3a78b42ab0a757e0d0eca9ca61"} {"id": "train_046196", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["governance", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "520ed50d1efc8ec869efdd34b9565a924b25281d"} {"id": "train_046197", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["research", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "74b0b01cbf21e872afcb3ed0b71a3926661f0b72"} {"id": "train_046198", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "688c5aa2d7b927fde112358a809e4c3ccdd8ec8e"} {"id": "train_046199", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "f544c340623b25bde89f8ffa605c50e6411fbfa5"} {"id": "train_046200", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "dffe0bbc840323f93990d3a33ceb53d8759e8a72"} {"id": "train_046201", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["security", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "b58e7901b6e9ccdf331180dd4708ee9f2c970628"} {"id": "train_046202", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "4ef1378c01f5b5cc43728e58794a72b31b3811e3"} {"id": "train_046203", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["security", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "eb6e9340c1b452b4a0bb96068ed4bca5577c6c77"} {"id": "train_046204", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "a2332ddbdd02ed104a1786fc82f414f7dc393f39"} {"id": "train_046205", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "4b75b7e76c6f21dac282e6fff409e433463c38a3"} {"id": "train_046206", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["data", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "a7a4e693607f6df321c58cf4b33d60116debf4e7"} {"id": "train_046207", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["governance", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "c8d94c96b8b55711f8d0fc248c66edcebeb9cdf5"} {"id": "train_046208", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["data", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "957cbc3e7158e9f6f3f4b1dd35970996d0849831"} {"id": "train_046209", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 9,\n \"confidence\": 0.9,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "fbb966f9ff51b472bf4643479040a350461514cf"} {"id": "train_046210", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["security", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "818076166014f1761ef7f03fbbe75b929def51f4"} {"id": "train_046211", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["research", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "861d402cd44118c07b3c6cd0145a725e7a2735eb"} {"id": "train_046212", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "SQL", "moe_experts": ["coding", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "4e8609f3f1b6f7ee11418b6dca3b40835b6195bd"} {"id": "train_046213", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Bash", "moe_experts": ["performance", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "6e43e4e8f086f211a2530e519bbff016075a4370"} {"id": "train_046214", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Python", "moe_experts": ["agent", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "96fb2126997aaaf861a0db370e83e798d1339f0c"} {"id": "train_046215", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "fc25ac099ca089f646109e2a094a067dd3b89e84"} {"id": "train_046216", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["security", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "ae533309dc1bbed3658f0c06c66902619bf0bca3"} {"id": "train_046217", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Java", "moe_experts": ["research", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "a634aee7a9804c5a19f310e235f4891e81111b88"} {"id": "train_046218", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["performance", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "c8e3263567e3bf6deb08a8bf9fae6636209937ca"} {"id": "train_046219", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "e58b4795640a7938203895d5ed5250d6add4e866"} {"id": "train_046220", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["security", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "2a10bc6a89c998d3282bdc56d110ad0b4bcf1781"} {"id": "train_046221", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["coding", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "8b136055fd5ab956b2de686c828a931147e63a11"} {"id": "train_046222", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["performance", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "4405366d1e1e1eb2f19385f7e9d19720062fdefc"} {"id": "train_046223", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["research", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "5f1d09dd5e6e42939817e2c357e589fd3646e260"} {"id": "train_046224", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "f57dda6bf8e6d0394bbcd51879a3914d42060d6a"} {"id": "train_046225", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "b63a538ff9a672e28f6465cbe89052f2afe1e828"} {"id": "train_046226", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "8db5c0d3d4bc956a560e6b865cbd7e9dc8958219"} {"id": "train_046227", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "0afab87473e921d1eb85dec68555c0a3b807f5bd"} {"id": "train_046228", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 9,\n \"confidence\": 0.9,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["data", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "3bff110fbe6cc23c035da245d561a1595f11e5da"} {"id": "train_046229", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["data", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "159d9c89bb08bf18de9038654e3df166b6e7c05a"} {"id": "train_046230", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "f7a9453046b3de095a11dc9f6324bfda03084f4a"} {"id": "train_046231", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["data", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "f75c69849d56573e691e7130371f5a415df02533"} {"id": "train_046232", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["coding", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "7d7aa0c69a3d6dac9327ec84279400fe057d7a96"} {"id": "train_046233", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["data", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "4012fb4a145746422fa4a20e1974b6999e444a45"} {"id": "train_046234", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["performance", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "61e6eb48b0ec3f7d768c010fb4db0a19d8d1f1bf"} {"id": "train_046235", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["data", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "0b8038a058ba082e1e1f3eeb015b621ecbab0083"} {"id": "train_046236", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["coding", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "d128a99853eb78d08eb6621ac187811480dd7dd0"} {"id": "train_046237", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "2a82776090fc58a28d5e89b775696f3b102fd4da"} {"id": "train_046238", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "b8eff0eab40e3dd459706ca240e6912cba93fac7"} {"id": "train_046239", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["evaluation", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "2b4475f87bc9b5e707dfdbcfbe3fc867d9e981f0"} {"id": "train_046240", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["research", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "bdf6dba2cf9b0feb041fff864b1ddf415b9f12ab"} {"id": "train_046241", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "C#", "moe_experts": ["data", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "20802ba09e303a1edcc8a701cdbc874220caf724"} {"id": "train_046242", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "28bd9185e71d50e2522f87dc27dea62eb591141d"} {"id": "train_046243", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["performance", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "f16cf988ac4d90930177611eb0c17d867125cd42"} {"id": "train_046244", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Python", "moe_experts": ["agent", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "7f92eccf8f911441e21960955dac028e32d36610"} {"id": "train_046245", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Python", "moe_experts": ["governance", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "5b0b84538184c6170515ab576f703a4fa27afc11"} {"id": "train_046246", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["research", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "ed3c000005998df0d397a64c9463619e6e2e345e"} {"id": "train_046247", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["coding", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "689e9eaaa3c65ae094f0f7def4fea51451f0a901"} {"id": "train_046248", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["security", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "26c9d6f9b65665e50b9b89854473fd70491f959e"} {"id": "train_046249", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["agent", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "6443d6bae5ce140a2103b990b37c7f6a9f8ff127"} {"id": "train_046250", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "40041d5621642c2b3e6e702aeefee666be576739"} {"id": "train_046251", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["coding", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "94a80f854359d3e8e81848eeee486133f5a794f4"} {"id": "train_046252", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Rust", "moe_experts": ["research", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "e4f2e6e90b842844879f6c7c013af48f75bee039"} {"id": "train_046253", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["governance", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "e798bb5cab8fc8982cd4e49d5ec8bf0cd51b4e56"} {"id": "train_046254", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["performance", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "57a828923fae4c41de42fcbda96e0a154361cca6"} {"id": "train_046255", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["performance", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "1e6212cca8ebc0dd9825734542583ff79baa51e6"} {"id": "train_046256", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "88bb42dca313ccc79d6c5b3a7c09accac018cf08"} {"id": "train_046257", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Java", "moe_experts": ["coding", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "a57a3d339da7e42a445e7c8984c817c65f3d63f6"} {"id": "train_046258", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["security", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "ab3bb3efb66d162c0115e4ebe630cdc8d40accdb"} {"id": "train_046259", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["agent", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "8e8c15786b04d4b864e506efa99b8ea3379c818f"} {"id": "train_046260", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "6908066078580ab42b46f19816605708e3973eb0"} {"id": "train_046261", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["research", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "6222e6fc2103e6803c3561b9da90ad04c0df2b37"} {"id": "train_046262", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n cost = 0\n for _ in range(budget):\n p = plan(issue)\n patch = edit(issue, p)\n ok, _ = test(patch)\n cost += 1\n if ok:\n return patch, cost\n raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["data", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "9e7f1853b98c5000e5cc82f1dc82300351f048d4"} {"id": "train_046263", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["security", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "4333aef856b9d7614b87b1fe2f2f5e90b9b4108b"} {"id": "train_046264", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["performance", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "ec9eebfeb3cbca4c1554aa47dacb181dc83493d8"} {"id": "train_046265", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "9dabca7400060565758de54a1433cda8fb7ea4a3"} {"id": "train_046266", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["performance", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "373d73e93b56f879c986ac69999dc3e6b39b8f7d"} {"id": "train_046267", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n cost = 0\n for _ in range(budget):\n p = plan(issue)\n patch = edit(issue, p)\n ok, _ = test(patch)\n cost += 1\n if ok:\n return patch, cost\n raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["performance", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "9e7f1853b98c5000e5cc82f1dc82300351f048d4"} {"id": "train_046268", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 9,\n \"confidence\": 0.9,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["performance", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "25401b610b1e05266d0a1a73aa649c60a750b407"} {"id": "train_046269", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["coding", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "5627330a3d3f5cfb1a2b39e9c64602012d19d019"} {"id": "train_046270", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "2ed6c5b82b7769d260909fe92a3e940395e20e6e"} {"id": "train_046271", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "8c8417891056429d3e05b7932880289c72e2012c"} {"id": "train_046272", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n cost = 0\n for _ in range(budget):\n p = plan(issue)\n patch = edit(issue, p)\n ok, _ = test(patch)\n cost += 1\n if ok:\n return patch, cost\n raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["coding", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "b236b94e2fbd782daced73facba133228d235622"} {"id": "train_046273", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["security", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "33e8e29742b8c3e387e422c291ace1c765fb13cd"} {"id": "train_046274", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "95417d2b6cc948e7393e0c32956ad6df2793c1b8"} {"id": "train_046275", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["coding", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "f063217e799d1d2b84c270354ee4f567b57fd5cf"} {"id": "train_046276", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "984aad36d5a9dabe5b6c229892c5d82e1243226b"} {"id": "train_046277", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["research", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "3b5d05c6f522d7bce1901ab151f62ca6f9f39fd3"} {"id": "train_046278", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "6b118f34f3a85ecebfd0c434743054d00eb56a77"} {"id": "train_046279", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "47b3d8035fcd5955ac954f3d1fb157c32b0a7ba6"} {"id": "train_046280", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["coding", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "8b136055fd5ab956b2de686c828a931147e63a11"} {"id": "train_046281", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "bbe6609549477db1204b5f9b847dbed762ec4fff"} {"id": "train_046282", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["security", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "9c896f532384e5f56aaa5f032fc89f6a9b27cbcb"} {"id": "train_046283", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["research", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "72bcb681f614f2f27998de4881972dedd96614a3"} {"id": "train_046284", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["security", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "055f3c3f079724af7ff18b4914537f7ddb1d67bc"} {"id": "train_046285", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "1793d73fd443d51589cf6f63c1731efaa37b55bc"} {"id": "train_046286", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["performance", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "d50e81d09f4c13cf8f66b3611ddd553cbfcc9697"} {"id": "train_046287", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["coding", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "52cad26c790c8a3e96c4f1a5ce94b49783c79e72"} {"id": "train_046288", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Bash", "moe_experts": ["research", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "43b52982b67ac44cb47daefe6abca95a1eb5a4d4"} {"id": "train_046289", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["governance", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "f8298d41c50401d373cc476adf9f1edfcf28353f"} {"id": "train_046290", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["research", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "99f9132a59cc94b85df1190711ee200bda9096db"} {"id": "train_046291", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["performance", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "235a796ed628f47db4684d117e3f0115e916cbd3"} {"id": "train_046292", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "a3a58f30ca68e9e5a9c6a401f8d3b4c4e90919f8"} {"id": "train_046293", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["research", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "efcada577410cf93b7ad13b422a8146091afda56"} {"id": "train_046294", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["governance", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "2fc5a81faf30fbdd24c1436ac9e00e4a98199712"} {"id": "train_046295", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "45c3505a601c417305a8222c482829d4289e1273"} {"id": "train_046296", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "780d1dc06001b3ed3f0e77f10f99ada638bc2f69"} {"id": "train_046297", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "101231df9d3a763bc8aae35115d1e3fc6c3be120"} {"id": "train_046298", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Python", "moe_experts": ["agent", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "840112f24b2bc1cc3299f4dd6614a0fa8ad49abd"} {"id": "train_046299", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "05ac77711c79499ae41fb177f1d018a13563fa42"} {"id": "train_046300", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["evaluation", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "b1643655b1d4cbc2ec5f90924529319ff53745ac"} {"id": "train_046301", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["agent", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "d301d612be5e1d1d5589318fcbad81efb2184f9f"} {"id": "train_046302", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["research", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "568a818c88c1fc9a116acbc6c19a00dae5319461"} {"id": "train_046303", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "b038dcafb689fe369f36632db9e7dfa883dce6c1"} {"id": "train_046304", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["agent", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "6bfa923961ebf79057ee39097797c66a016f3a84"} {"id": "train_046305", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "SQL", "moe_experts": ["security", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "b20b0c2f2285741aa1532713764ffc5cab1a9a28"} {"id": "train_046306", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "9abb640b30e77b90a7efa637fd510c085165e790"} {"id": "train_046307", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "1e76f42cdb5fcc5eb9b2a8dcc890a9e0df36882f"} {"id": "train_046308", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "306e4d3f6fda09e0beebe05b0306567f7b464ac0"} {"id": "train_046309", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "41151031f3a7e6633d502a08214322bf64bda25c"} {"id": "train_046310", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "e5b2816267c1aae95aba6aeb48cf6d96f8645fc2"} {"id": "train_046311", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "98d813860002845d670e83d810902f45953009d0"} {"id": "train_046312", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["data", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "d8c01d1463bc5100c4d068e5f32b6d989d545e4d"} {"id": "train_046313", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["data", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "85b53ec95bc2e085e26f2663a32a75c683f7b379"} {"id": "train_046314", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["agent", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "2ae91c85f768f0a080e68b9d3a40ab35b0da7fff"} {"id": "train_046315", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "5f841cf58f591ed88247b70fdcddb68f220b22f2"} {"id": "train_046316", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["governance", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "08f902af149364d9206cad1eb4ba10d1bfa27e08"} {"id": "train_046317", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "fb24dbdcb44a0e24f2abc7e3089a2fa585096ff5"} {"id": "train_046318", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "566a16579aff4b7026b04ca236b2bac1ffed681c"} {"id": "train_046319", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "b74008756070bc01c419693967ebe20f0893c4a2"} {"id": "train_046320", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "e0d1bf5c90a5512128bbd45d4764ee3dbdf36e8e"} {"id": "train_046321", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["security", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "f3de738f25dc3344031eef24e4c6f594024a5f47"} {"id": "train_046322", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "c0b3bb456077b3e56a54d57ff101ff3beb6a9455"} {"id": "train_046323", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "10061fe6fc31177307aeca0f0bef2020192a595d"} {"id": "train_046324", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["agent", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "e09bcf3021369e934c2383f6084b9061e8971fa6"} {"id": "train_046325", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "f52a71e1c3de1f3e2344929b674181cd2a0e9952"} {"id": "train_046326", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["security", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "c85f6326d644e1cd631a5c1a6e180ad160a24729"} {"id": "train_046327", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "277189aeea628e93abb491910e2020c1812f4909"} {"id": "train_046328", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "313a79c02ab96e81b338bd132d22ad4e3818dc7d"} {"id": "train_046329", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "98ee1ab40bd488279adb58857574a747908241c4"} {"id": "train_046330", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["research", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "4039f3000888b2c29a86d3762c39f2d6e5183349"} {"id": "train_046331", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "2de24f3018c2218734bca261c99a7f5bbc3af0e3"} {"id": "train_046332", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Python", "moe_experts": ["evaluation", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "e6075bf7f628a8743b187b39aa0d880246234395"} {"id": "train_046333", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "4b209c59cef8a3fcfc5097c9ba887d88fdc95073"} {"id": "train_046334", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "c095edccfd191e392c70ec4fe01f75825f2e594c"} {"id": "train_046335", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["data", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "0b4e567df5943661e75965b1b4535c1e37775cb7"} {"id": "train_046336", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "a7801cbc551493225964f5212398a6e7f31a6a69"} {"id": "train_046337", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "af9fb3c897591526de2fba8eedfd64dcf75479e4"} {"id": "train_046338", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n cost = 0\n for _ in range(budget):\n p = plan(issue)\n patch = edit(issue, p)\n ok, _ = test(patch)\n cost += 1\n if ok:\n return patch, cost\n raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["performance", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "fb9ca4d5b9fe8574655ea304f0263fca5af65c7c"} {"id": "train_046339", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "fba64e76bdf389629390efef560ca74d786618ec"} {"id": "train_046340", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "1abf0c9657b4d32d82a9e1fb12b7eb46a8ea9b1f"} {"id": "train_046341", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "417ddb70a9166954ecfe95c87a8df123586cdb23"} {"id": "train_046342", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["data", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "023a8924ab224fd020fd893fb0ee7fa7e1ea2e91"} {"id": "train_046343", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "900bc418a2e20d7a645f200257cf82cea0d4a771"} {"id": "train_046344", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "2aad973126e5a000e2ba0a7844228b17d6d03f7d"} {"id": "train_046345", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["research", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "0048dcbaa676c76d514fab1a4d1a32ec4da9bf3c"} {"id": "train_046346", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["security", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "4c2d7e98302e1a276b7ab725b2627481d2b6f1be"} {"id": "train_046347", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["research", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "703dd8161c879658aeadff332ebebfc45afd3449"} {"id": "train_046348", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["agent", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "ca3c203bcb021f1b99acb9058868b4e20912f91d"} {"id": "train_046349", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "e0a33d8b7c16b7f2a206f8cdd15da982638e85f7"} {"id": "train_046350", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["data", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "fa0dbd5f7c8cd011afa50a44e0bfa9411189dd25"} {"id": "train_046351", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Python", "moe_experts": ["coding", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "0853c549da155cc1ee15670f2032ff06f3306698"} {"id": "train_046352", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "2545ccf251a15b4fac1a9737916c9d9e2214308e"} {"id": "train_046353", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "dd541bad414c2deec905394e71059edfad47d48a"} {"id": "train_046354", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Rust", "moe_experts": ["research", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "952a674bf79773494fb6d5c43208aa305aaf8554"} {"id": "train_046355", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["security", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "64ffa0aacaacb551e7d303cd9bce4a478f278d23"} {"id": "train_046356", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "2c76d9130780a6098ccf01e509937d07e632fa4d"} {"id": "train_046357", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["coding", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "3f8b62c0313c498eb3a8b3b33c343b7591a1f1bd"} {"id": "train_046358", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "20872573475dc6041dea1cb49636fda9bae4dcb4"} {"id": "train_046359", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "a4ff5190675f67e1f87049fce114d4938184eed3"} {"id": "train_046360", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "c0d06ab8491030059be2e5d4410b93ab7efc0301"} {"id": "train_046361", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "b0ff43a76af0ae35d583434edf65931423bcce4f"} {"id": "train_046362", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["research", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "4112a719f6d30822e304ec3f5ade8bdf49a822b1"} {"id": "train_046363", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["governance", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "1d149bad816299e9dd3b3a283e02c1f47829a5eb"} {"id": "train_046364", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "4cb2d692e0ce6b5b277d42ecb091a5facf172bc0"} {"id": "train_046365", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Java", "moe_experts": ["coding", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "4bb200a8f7d27b39a74dec9850416072c07bd302"} {"id": "train_046366", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["security", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "7e0237cac5029a978698e40ed68ea658296d0644"} {"id": "train_046367", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["data", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "c68baa5c815e414d931195110f7c2b16fb086daa"} {"id": "train_046368", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "65545a6514b51d400727a600c79a40bcc0e7454f"} {"id": "train_046369", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["data", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "e12a77bb0d33bbba92889d5215f5fbc8290300ef"} {"id": "train_046370", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["performance", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "3351845424ef58085749162768503f51829f7582"} {"id": "train_046371", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "SQL", "moe_experts": ["coding", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "db1ddc4708db4aa52a4ed8dde66deb822f4e82b0"} {"id": "train_046372", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "0c643402afa1e78966b6a673a5c2e3c77a561142"} {"id": "train_046373", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "03cd4a256a075b71a6bc0a41cc578de672f7b815"} {"id": "train_046374", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["agent", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "68ac3113d6cdb75e06a2dec120b5696f5bbe4d68"} {"id": "train_046375", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["agent", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "492811323e714392e9c38ca8db272fe3a32cc9b4"} {"id": "train_046376", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n cost = 0\n for _ in range(budget):\n p = plan(issue)\n patch = edit(issue, p)\n ok, _ = test(patch)\n cost += 1\n if ok:\n return patch, cost\n raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["security", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "39b2d8bc9c08330dcb60efc0c2a7050b66e7c6c4"} {"id": "train_046377", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "27aaad0cd7a8f0819b3e5842a388798c5fff5c39"} {"id": "train_046378", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Bash", "moe_experts": ["security", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "480e6a0ed34acf5b711998d18062b5720542b31a"} {"id": "train_046379", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 9,\n \"confidence\": 0.9,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "e67e433248a01323d3fe2b0fdb245451f5f39fae"} {"id": "train_046380", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["coding", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "e0100b1a6c246288a06a170f332ba469d1e985ca"} {"id": "train_046381", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Java", "moe_experts": ["governance", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "7bad13b717b936616e9ee71ed2dae747294d1040"} {"id": "train_046382", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "C#", "moe_experts": ["coding", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "6e75f3863e6124c7fc597bba01b1025e39b4fba7"} {"id": "train_046383", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "cee3b7f2e0bba2e54f65278fc16f3fd0d71ebae8"} {"id": "train_046384", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["data", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "a4eafd372d8c35db5f7ca9ffa2ae3de8d7e5cb40"} {"id": "train_046385", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "6d2d74ae1cd0237fba4111a9b4e4aaee4be59734"} {"id": "train_046386", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "26e8c01c43f4895dbef441014ab31d01bc8f8e58"} {"id": "train_046387", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "f20a782714259dd22f754846cd337d5055b86d19"} {"id": "train_046388", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "61d8e4f042e6b467bb6249a155d9d3017919e6cf"} {"id": "train_046389", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "adc71ecb3d1c724fc2c75c9504dbea357bc64d20"} {"id": "train_046390", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["agent", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "32ac6d1d5af6dfe11241662272d8f40840b01892"} {"id": "train_046391", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["coding", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "2bb73f30aa8ee1c45e9b73c2f8fce08629db35b3"} {"id": "train_046392", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "a054ecf33b1329420e6809fce079859eda4fb039"} {"id": "train_046393", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "6e3d896febab80cd934c838dfbebe420c63b94bf"} {"id": "train_046394", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["evaluation", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "c1c05ce98280cada21015e02c96f605100365c63"} {"id": "train_046395", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["security", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "e70f9cd4bd300a4a50f84292c8e78c96769bb4e7"} {"id": "train_046396", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["coding", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "e46449e4a4df9b7c405455f95c6e1afa4b1461de"} {"id": "train_046397", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "60216b7d7633194db4caa9cd1a52cd3552e5d9d9"} {"id": "train_046398", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["evaluation", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "9cf4d30c1ef3fc110093b8a77e95ff94f5a6870e"} {"id": "train_046399", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["security", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "1e95a7ad23bdbd53b53122276acd98165beeb676"} {"id": "train_046400", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["coding", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "0f3ea715a0118f4527a29661e7646c503e55ca3d"} {"id": "train_046401", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["research", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "594b2e35fbc49b6b223ed1807cd0c8f35513b64a"} {"id": "train_046402", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["evaluation", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "b99443d39edc9d8d6e9067341a9e4571e40c53b1"} {"id": "train_046403", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["agent", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "9c781d0850d7e6eff9c47eaf1ccd5a90c379812d"} {"id": "train_046404", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "326ce3eadde28ebb02d4df33eec2d8ce1d962344"} {"id": "train_046405", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["performance", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "f134244e21860071018f452390be6bd83b9fd644"} {"id": "train_046406", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["agent", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "4dd9d29af9db9802a6f57a40d8fcea1c7ca431eb"} {"id": "train_046407", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "e4b5f4ba831f22caa16135baa79354ed14693276"} {"id": "train_046408", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["agent", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "1217143823211c7e0d39d20864c4ec9c21f92af9"} {"id": "train_046409", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["data", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "a203addcc62fedea8ff034152f8876dc43fb05f1"} {"id": "train_046410", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "cd2f0bde2cfe9a3c01ebcc0cfe4ff3387b3cfd7d"} {"id": "train_046411", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["data", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "278081b53c4fe663dc6360a6f962d8fbb217e911"} {"id": "train_046412", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["data", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "85c490cf7c3f0e5c326bec8f56b618fe526a6966"} {"id": "train_046413", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["agent", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "a59696228bbf22f740dd4f550c85817fb4c5564d"} {"id": "train_046414", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "8fb0a33b3879fcc4d65c562a5f7a8890e85e7092"} {"id": "train_046415", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "47d5e3924f1613372e402987887d81de1499ef74"} {"id": "train_046416", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "630d8f949160adbfbb6ef4e9861bd8d47f210986"} {"id": "train_046417", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "C#", "moe_experts": ["performance", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "af4ea35e355aa9f030030bf170832e82c192c038"} {"id": "train_046418", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "94b400a3807f9a8d016e0dbe3bb110529edb634c"} {"id": "train_046419", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "f30d16c4c6b23bec248b600c4057e5afa74383b4"} {"id": "train_046420", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["research", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "d4cc4fd29f8ab7dba7b501d978529e0551836759"} {"id": "train_046421", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "215d53c301145bbb5e8bc95b0ef449048f068e7f"} {"id": "train_046422", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "5933db2c1abbbfe5d4f491decdad2ccdd290243c"} {"id": "train_046423", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["data", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "370c5d89fb1c8540f552c64daeea0d35fe68e0ea"} {"id": "train_046424", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["research", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "b2c3e14685789c15aa76906e95d44a013828b089"} {"id": "train_046425", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "abc5f2b11ca7e71cd6a71bfdd8d0af57c814dd9a"} {"id": "train_046426", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "64d591c19c3cc678791fc7c237625cbc4707963c"} {"id": "train_046427", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "0ced0b59a01e204ba1063482060b29b7eb06d257"} {"id": "train_046428", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "85d730c43b7106a11342bd1db02a945249d5a671"} {"id": "train_046429", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["security", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "f5e8408d43b1599a7b99ee271c5a08aab0f12053"} {"id": "train_046430", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["data", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "956f9ccdd45312c30f2bacbee54498667a9bb939"} {"id": "train_046431", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "C#", "moe_experts": ["research", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "648218a315ad11a991c55b8c7bf59b6febb41132"} {"id": "train_046432", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["research", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "737cf0ae20f4e09d0b5a950b34033f0cf68e1c54"} {"id": "train_046433", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "27dfd2ab45d5f0ad233f564a675853eb33f080fc"} {"id": "train_046434", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["evaluation", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "760981963b09cfe791bff811799f942a1b508718"} {"id": "train_046435", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["security", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "f7cbf25e96a860ebb69b6e393ea6019835f12f82"} {"id": "train_046436", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "4c8cda7d22abce6c24f28a478738bb9f786fd046"} {"id": "train_046437", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["security", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "01ff15273612e656ab2cd137e67ad7ca475da217"} {"id": "train_046438", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["coding", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "be7a4e86924ecce35e1e17dcc25bbcbd6c5dfb41"} {"id": "train_046439", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["agent", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "443c76948aac1b5ec75fa86ec835fa1c56f82bd8"} {"id": "train_046440", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Java", "moe_experts": ["governance", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "165047ffa0789b605c5b29cfd9eaf6dadf34ffcc"} {"id": "train_046441", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "cae203a644bff85cc02d6fa581d1911bb8b70b38"} {"id": "train_046442", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Python", "moe_experts": ["coding", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "a4acca85eab2641f31cfeaedffcafa20a1f722da"} {"id": "train_046443", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "af16ba7d44f512e50e7adead691785ead5247bf6"} {"id": "train_046444", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "d0e4f7885824c5f666b8e2d52753ca12111d8368"} {"id": "train_046445", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "c94133fa045653ba30ed2ef4ad026c41fede1878"} {"id": "train_046446", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["coding", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "a9c6f8e886c25611999e48ecdf62bd0583aa54ff"} {"id": "train_046447", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "C#", "moe_experts": ["data", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "11500a44c573016cdedf9cdbdc2e4df5036d28ec"} {"id": "train_046448", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "e9609266339b0ce700a0993f60a2facc9f98ea8a"} {"id": "train_046449", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Java", "moe_experts": ["coding", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "4a143e148e6efff664b730f731f536e619dad61c"} {"id": "train_046450", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "85558d4201e13e92ddd44089053da4cdf91d91dd"} {"id": "train_046451", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Rust", "moe_experts": ["performance", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "3c37768fd35fd837d816ff4c0be25f10911d39d6"} {"id": "train_046452", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "d8e5c3334c94fc98d1b6581c1dc0645ae4f66c3d"} {"id": "train_046453", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "964c2c3bc25a751327c15380010a1303a7834ad5"} {"id": "train_046454", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n cost = 0\n for _ in range(budget):\n p = plan(issue)\n patch = edit(issue, p)\n ok, _ = test(patch)\n cost += 1\n if ok:\n return patch, cost\n raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["governance", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "dc7980699929340416d490334df26a8227fab3d1"} {"id": "train_046455", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "75b55a5fa591b1ef59b2f6a3db6982bbd79459aa"} {"id": "train_046456", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "d9572af470a63422049e8b0168e8cafd8d3f85eb"} {"id": "train_046457", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["coding", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "72647c8f15b2ea03de0a5392df36f424af76e626"} {"id": "train_046458", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "6538a4711c511898af65193f06810b2dc426ff22"} {"id": "train_046459", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "245d0ff7c5ee8444d6e01fc6eaf3dec257f7536c"} {"id": "train_046460", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "852aba3fc5ec5dc2406ada111a93248dc9ade8ea"} {"id": "train_046461", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["data", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "9e9e07f06cd54b1a9322f7fdcc192a87088710ab"} {"id": "train_046462", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "e2f97fe4338012755ef1d4ee0b52be89c49ef3a1"} {"id": "train_046463", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Python", "moe_experts": ["evaluation", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "f13109ff9507483bdcffdc5d5cd5cc56fa683e4f"} {"id": "train_046464", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["evaluation", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "3edf065c5a71b72f7b4718d573cf39c33fd8d0e7"} {"id": "train_046465", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "4666dae49c5f2868a0c872312b96419b673491fa"} {"id": "train_046466", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "SQL", "moe_experts": ["coding", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "7f29cac5703b8e51f56f140e458a7da405208472"} {"id": "train_046467", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["agent", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "c0b084c12c1637a6c5b3352074dbec491b62192e"} {"id": "train_046468", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "SQL", "moe_experts": ["security", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "4c71a0ba6249a29a2767bb61ad4e46b6b4a7f82e"} {"id": "train_046469", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "026eaab5afdde359f87ed0fc45776c3186192290"} {"id": "train_046470", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["agent", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "97850ce13fa0a9b29b21a8c328eef72b1b192dc0"} {"id": "train_046471", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "e87991cf3a4270ccf16f5894b63bdd91f05239d6"} {"id": "train_046472", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["governance", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "4a773d76b67466d487f0590c32212a2f834bde68"} {"id": "train_046473", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "89e92c9f3f4a0d0f99b25edd53b298701ba158ad"} {"id": "train_046474", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "b57b626e0ffa79d05143208acb65571dc6fafff8"} {"id": "train_046475", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["security", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "041c6c4ff8d37a8fabd03be8439828baaf2f5950"} {"id": "train_046476", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Python", "moe_experts": ["agent", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "e94fcc5c4d24f400ed7f452074479d2267780f84"} {"id": "train_046477", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "d6ce6847e787e7be9a94d7bbcae46df958e11533"} {"id": "train_046478", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["coding", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "197a2637e9591db13ecf97bb90a35d26c11e242b"} {"id": "train_046479", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["security", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "e0eb024413b6779bc73e3f37a03aa74298fe908c"} {"id": "train_046480", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "456b73c6af99d58e8422803b270bcfaaba1a9c97"} {"id": "train_046481", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "9e254152214d1b7f2bab0b82e9c1207d0d9f9a23"} {"id": "train_046482", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["performance", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "ea24bcff2e0f6fd47a0b842dc7328755fae529bb"} {"id": "train_046483", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["coding", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "7d39dc9ee7905865d0bdf9b136b569ae91683507"} {"id": "train_046484", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "e14cf0380b62cbd8b854ec76d1cb7631db0566db"} {"id": "train_046485", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Java", "moe_experts": ["research", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "eae4f490e8252baaba37a26a6c6e6bf8642fbc9d"} {"id": "train_046486", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "e6c88962efdc6072f3f9eaed9f710a23a968e9c2"} {"id": "train_046487", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "8739a6c0a52172b57d759cd0616a3c84ceee36c5"} {"id": "train_046488", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["research", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "e32a8ff945bb20c4cd923eb6bb8987632108e502"} {"id": "train_046489", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["research", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "a1a028e9eeccac8893bc7cfe0a4bade88edaff83"} {"id": "train_046490", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["agent", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "0cdb1938dfabd976ef62523f3fea6d8e3e35ea00"} {"id": "train_046491", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "6d3872d7afaf8fbcf82b98d8ba7a77ee63024e4c"} {"id": "train_046492", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "47ae8f2f6dca3c26e96d7d951cedf982e08129b6"} {"id": "train_046493", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "fd4c6afbc6c48b854697e06e8be1d75f013701c3"} {"id": "train_046494", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["security", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "c9d5ae9a4f411758424c7d0e801411f29595c09e"} {"id": "train_046495", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["research", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "81a48c1bf2d9c068f47104dd98f58f9f0730e1c5"} {"id": "train_046496", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "9198756a2d78b09120004c30a08114149954744b"} {"id": "train_046497", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "e3a8dfe81a22afbcb2edb1f43a739b32a211622c"} {"id": "train_046498", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["governance", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "7ce85d944f478e46287b4be3146a9bf927169f2d"} {"id": "train_046499", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "173d54ed4e75f02d6383930bdaa2104f9b4316e3"} {"id": "train_046500", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "5ae88b099d66bfeb972058e7f10bb799de77fcac"} {"id": "train_046501", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["security", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "b5b27fb1e8178d21e7c911351ea7322b032d526d"} {"id": "train_046502", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["performance", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "2ae91c85f768f0a080e68b9d3a40ab35b0da7fff"} {"id": "train_046503", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "58aaf0d8932d94daac16196fcba768ac88d60747"} {"id": "train_046504", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "f6f8a13198636eb120a52ae2366df4ccc002c2d4"} {"id": "train_046505", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "29aa510ef40a10ffdc91f65faf20dad16e6b8f5c"} {"id": "train_046506", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["coding", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "dbe281be2a9cbe561830d01fbbf6d0b5b28c899b"} {"id": "train_046507", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Python", "moe_experts": ["coding", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "ef62059cecc5e08a20e657a36dddf24a7e446fbf"} {"id": "train_046508", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "944ce600ddf06a66c1ab7947e9642363d359bcfa"} {"id": "train_046509", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["data", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "6e3c7540f2119046791ed4162c5c289502c8c2e2"} {"id": "train_046510", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["research", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "dd3c38e342c1a18018116932a5ae585dc8cf0cb8"} {"id": "train_046511", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "4b1475cace1af4abed8cf0cbe39501c3e2a04fbb"} {"id": "train_046512", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["agent", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "3676a82b5282446d0bad16da239064030cf0c56f"} {"id": "train_046513", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["research", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "c1a6776d212e63f37f02443f57ca56d707946987"} {"id": "train_046514", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["governance", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "99f3e5c4876fe240e2502cd662b2a360f77d9f3c"} {"id": "train_046515", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "106190dae60a369bc42dd1d832c658d5512df665"} {"id": "train_046516", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "4112a719f6d30822e304ec3f5ade8bdf49a822b1"} {"id": "train_046517", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Bash", "moe_experts": ["data", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "142a10813380819e812e04d4ca5d4de0feb2266f"} {"id": "train_046518", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "61a752c1d8ddbeff2a3dff6782554ce44137b0b9"} {"id": "train_046519", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "d33c668ceed64b081893f73d71bf7b1408bc1e41"} {"id": "train_046520", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "SQL", "moe_experts": ["security", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "730a9a055557999f12d38621d506dd69dc95f7c8"} {"id": "train_046521", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "bb88e7b7d2712593e75b3f3c58d553da2032b8f7"} {"id": "train_046522", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "af9fb3c897591526de2fba8eedfd64dcf75479e4"} {"id": "train_046523", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["data", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "d9564a41b68a8ebc0108a27f850322bc421eebb6"} {"id": "train_046524", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["performance", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "ef73773ce8d321ef20d1f45370f456eef317240a"} {"id": "train_046525", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "a7e608f36ec16775b5d642df0b37eaaf9fcd1756"} {"id": "train_046526", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Go", "moe_experts": ["governance", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "88019c0aa91943fb8a261e8a69f3d458e17fe7be"} {"id": "train_046527", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["coding", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "11ca797af81f1a7944c6908e44d08117fd21ece1"} {"id": "train_046528", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["data", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "80e4ec24c5f14fae16dcb07154250d13651be5bd"} {"id": "train_046529", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Python", "moe_experts": ["data", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "a99062cc419d9b5a0c44f59914cfe70eded945a6"} {"id": "train_046530", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "SQL", "moe_experts": ["research", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "e2f2bbbe4feb4056f587bf5506c05c723135692b"} {"id": "train_046531", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "c9e41e37bf313e1e5a7a2e0fce535600f242897f"} {"id": "train_046532", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "16847abfe64f486c78f220ab27ce7d998a5565e6"} {"id": "train_046533", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "da66f6c91b274a0a491a189f4fe7c03a94b85d40"} {"id": "train_046534", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["evaluation", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "88bfbaa8f00de2ce699d8572e95a1b22d1b6ef7a"} {"id": "train_046535", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["governance", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "eaac2bbfc7919a75caefebe4f48ae921c5e61a2e"} {"id": "train_046536", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 9,\n \"confidence\": 0.9,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["security", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "932d6f56f2753bb3e9f54b01b157a8e939e9fa15"} {"id": "train_046537", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "164639556e0ef00b032ff763cd553395896fb91e"} {"id": "train_046538", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "008de7b5448f6062e3454869a5130a072f764f33"} {"id": "train_046539", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "d5fb23ef48cf2a42a30153397f9ccbe60da4e0e0"} {"id": "train_046540", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "6502d79da12be17c773583b437f8daf424e3aa55"} {"id": "train_046541", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["coding", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "f1c4f8054cc3d266d90052b91cdd621e6aaba9bd"} {"id": "train_046542", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["security", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "7fc3722ea6d9ca06722a657384ab8826b91d7a56"} {"id": "train_046543", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "e10814a7fd4c34319ff4cdf1335e9f5fa7ab89f6"} {"id": "train_046544", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "4e1f5821fc3d5dc19bb148e82b529072855ace33"} {"id": "train_046545", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["data", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "4f1141212e555e632fe924eedf289984a8556b53"} {"id": "train_046546", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "5190e4993a6e72e8198d916e2e134024748f9a0c"} {"id": "train_046547", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["coding", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "0ee23d9a15ebe5c17563768488a7629169d3bb9b"} {"id": "train_046548", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["performance", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "bd46bc4f68fb995eaec63cb832307b236b01485d"} {"id": "train_046549", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["governance", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "b0dbab00ffbb7cf157f61d33898d6b74d5099705"} {"id": "train_046550", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["data", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "4a1eb54c602015619f07d4397fa639978863546d"} {"id": "train_046551", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "ed80034e5246c4561290c800d09ff04e30e6cbb1"} {"id": "train_046552", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["agent", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "62848eb04203b9cfb926d5a35d81e1acbdd31465"} {"id": "train_046553", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["performance", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "6f0c3bc1ca64289979cd2de53254946033815e37"} {"id": "train_046554", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["agent", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "4328a1a63c10c79e111272e66ced67156de6999e"} {"id": "train_046555", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["coding", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "fdf87da2d0a66f82b4766676511756aa9d27f345"} {"id": "train_046556", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["agent", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "881bb2f97288cec4de7c5615eb6044a4f9c4142d"} {"id": "train_046557", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["coding", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "01ab69ba6b084f4a5165c969192a6f8087abbf36"} {"id": "train_046558", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "C#", "moe_experts": ["coding", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "331948a00c2a2261f37a3fa9a8c5f7de777c6086"} {"id": "train_046559", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["security", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "08b6e2b6cba0a05c503ee4359b4c1a3837193e44"} {"id": "train_046560", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "C#", "moe_experts": ["performance", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "240462aaf8fe71b56565a10386e39eea42511e19"} {"id": "train_046561", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["research", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "8a2fbde8cb6d3e4a13d2a06c37c8d59509dde24d"} {"id": "train_046562", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Go", "moe_experts": ["coding", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "07bf10f475bce9ab9be00c1fcda8da77f8583b47"} {"id": "train_046563", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "487dacefbb63ea5d54aa1911a5cd0eac4a13c814"} {"id": "train_046564", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n cost = 0\n for _ in range(budget):\n p = plan(issue)\n patch = edit(issue, p)\n ok, _ = test(patch)\n cost += 1\n if ok:\n return patch, cost\n raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["performance", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "9bd02ed242bdb4b71f47dd6d6eeef525489727a4"} {"id": "train_046565", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "b17ebbeb1e8f571cdc634d3e951a2329b5e3f0cb"} {"id": "train_046566", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Bash", "moe_experts": ["agent", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "41c7dd2f67df9f1900f32c5287bddc711e8b2f18"} {"id": "train_046567", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "b89df496a821566e69c0a91bbfad91f3e26f3a62"} {"id": "train_046568", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "8a052acbc1948a2016f497844583dfafb73f0cfa"} {"id": "train_046569", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["performance", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "b598e1e6383594470af7410ee6b86524ec254f89"} {"id": "train_046570", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Rust", "moe_experts": ["data", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "e3b55eec44929cf23b48272537c05e7f9d105097"} {"id": "train_046571", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "0be9997c572f7a723b7ef611d9c23d998687e798"} {"id": "train_046572", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["performance", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "5c918817dea20f00ca7a1242761ecb669ed97b46"} {"id": "train_046573", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "b09dbfadea18516cf79f76d345b9a2b0937d003b"} {"id": "train_046574", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["performance", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "b7eee55103bcb7aa1a1a945b41b799ac83e61615"} {"id": "train_046575", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "79b0b2db067d610291a71a765d1fe31659fe59b0"} {"id": "train_046576", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "SQL", "moe_experts": ["coding", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "16a8e4b4e7bb399bcaaa40113de6a8e297f8c493"} {"id": "train_046577", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "27fac08ca6a87e2180147fcd3e7c318dd0dfa02c"} {"id": "train_046578", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "848f7859bc35584c329b40229d002ac166f07c5b"} {"id": "train_046579", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Python", "moe_experts": ["security", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "d60bc1ea948006d53dbfba99ed1d7ac5002c02a7"} {"id": "train_046580", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "0340317aae39299daa5be761fc0ab926fbd1dcfb"} {"id": "train_046581", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["data", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "33826659dc6e0e5fafeac01234975aa0b8531d6f"} {"id": "train_046582", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "db361ef9a69f7ffbc0b43319dd1ba297bfa7471c"} {"id": "train_046583", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "9624f00b96a51342b0e792f5a133b47297bd6ecb"} {"id": "train_046584", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["research", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "003247c03aac727caa81619533398cc0ca24415c"} {"id": "train_046585", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "d0d328e5071eb73e44d7573a542ab5180d003337"} {"id": "train_046586", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["performance", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "9de99029518daf8c02cd6a125d8817e0da8a95a0"} {"id": "train_046587", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["data", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "11d52b4734631b0fde38231f4c06156a0f19df01"} {"id": "train_046588", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["governance", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "896528dc1333d278e3c196e444dece6acccbc2d0"} {"id": "train_046589", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "bf6bc6edb65c802d08c9f277d2db16db05f9540a"} {"id": "train_046590", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["performance", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "d9f377d764e6c5d50041727bd61f903d816184f7"} {"id": "train_046591", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "f6d0703bf02fa8492e14c4f768eac66af5172389"} {"id": "train_046592", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["performance", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "059a5ce9e7f4eccb31f99df1636a90b87c004961"} {"id": "train_046593", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "SQL", "moe_experts": ["coding", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "85361f59c8d351e71b4d137feda2b07648261732"} {"id": "train_046594", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "a37ad27321b6b1351200670ae663522bddb33609"} {"id": "train_046595", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "2932f612a5a364b603ff87b1be3ba7975d7eb5c7"} {"id": "train_046596", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "07f312f01f8f73c0dc6c4b93710f4cb6d4fe149e"} {"id": "train_046597", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "caa1155ba3ce5b5b7bce7c66cbafda6cb6d657ec"} {"id": "train_046598", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["coding", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "7a9d58a978a13b5e6b6cb728e46f8742ce1cf48c"} {"id": "train_046599", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Java", "moe_experts": ["research", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "9fd55bfefd2fd8482807c69bd7a706ccea5b05dc"} {"id": "train_046600", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "SQL", "moe_experts": ["data", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "ecfcc0d6308dfe93b33e87469985d20c79104782"} {"id": "train_046601", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["coding", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "759aa86a974fa2a7760536581c1ac053fe603dce"} {"id": "train_046602", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "c4355b5e6400a05eea1c1bae977b251ceaf7a2a0"} {"id": "train_046603", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "SQL", "moe_experts": ["research", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "d8c9192a59130103d0677a3b193074bbb89f9187"} {"id": "train_046604", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "c6953f7c6e2d7532e2458cfb457913ebf6334500"} {"id": "train_046605", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["research", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "228df8d949ab26de487db331eeed253c46856180"} {"id": "train_046606", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "07bd75c93a6e9c11bf0c8cc2f7d863d0712d23d8"} {"id": "train_046607", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["data", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "c036924b662584b64d5782207b2fd7553c2fe705"} {"id": "train_046608", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Python", "moe_experts": ["evaluation", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "a3544e72aa24f056321ee6590387667f21a38080"} {"id": "train_046609", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["data", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "c6953f7c6e2d7532e2458cfb457913ebf6334500"} {"id": "train_046610", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["data", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "e798bb5cab8fc8982cd4e49d5ec8bf0cd51b4e56"} {"id": "train_046611", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["agent", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "691c3160f40611f0aef4eedd837c0434dc568ba8"} {"id": "train_046612", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "2db594e7c9eade79fd645afe068b48ac050493d8"} {"id": "train_046613", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "edde062fe33371071b8ec5fa871ce2cc6d40ae7a"} {"id": "train_046614", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["evaluation", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "a4dba4051426a23c7d1703ff924b8885bb454976"} {"id": "train_046615", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "ea716d28335c4861bbcce4ec36b8f4a9ed99fe23"} {"id": "train_046616", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "d369183bd921d964406216d732765eb96d1ad85a"} {"id": "train_046617", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "f9f95a6633fa291cd0c4e873fdbfb8450eb479a7"} {"id": "train_046618", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["agent", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "9f7c0ef414418fd0156f3c0dab53d0acf7d893e9"} {"id": "train_046619", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Rust", "moe_experts": ["security", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "8165940eebc3729b93f5affc2a88393e91641712"} {"id": "train_046620", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["performance", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "df17eb44f22e47b99e0e2c806d83d0cca14db9c6"} {"id": "train_046621", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["data", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "729ebf6375c821d6e5bb93ebd51384b0e9009b26"} {"id": "train_046622", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "019cfde6172121ecf696dbb86c555a4f191b6d2c"} {"id": "train_046623", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "0a2ffd79e8d0a4a1dfa1a95d6731de6cf60eb2b7"} {"id": "train_046624", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Go", "moe_experts": ["performance", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "02c35d75a58f6b5e1f4a530da7c7325ef4bdd995"} {"id": "train_046625", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "6908066078580ab42b46f19816605708e3973eb0"} {"id": "train_046626", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "2a23d3003ba37c30d60ee3936bce435573cf7e70"} {"id": "train_046627", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Java", "moe_experts": ["performance", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "1687d8308f5426f659d5c20d850d1a9debdd3d16"} {"id": "train_046628", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "54eda9395a12fbe129276210cd728cb6b2bd0566"} {"id": "train_046629", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["research", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "aeb92c3dd1689b7444bb4b0dbc5bfa86401b23e9"} {"id": "train_046630", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["data", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "5f8dbe09a348cb772530d2f940ccfbc2107a2c9b"} {"id": "train_046631", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["performance", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "d983d5bc4fbcb49eed91fa560002abee14936af0"} {"id": "train_046632", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["governance", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "a42afc121f09f2419122a3c5b891e90a0261f22f"} {"id": "train_046633", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["security", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "cc5baa7c3fed2a159517cd5acaa8e76e9cb40a26"} {"id": "train_046634", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["governance", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "e3760cc363be39415e22d2b593a2a79deedd91d8"} {"id": "train_046635", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "8bfd8b327497b413be18411898dc965dbb99d432"} {"id": "train_046636", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["governance", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "79361a4eb73570765088b2aeca6b61c5a3c1e071"} {"id": "train_046637", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["data", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "5b3d597eb88cf9320f7de377351e0f4481f54149"} {"id": "train_046638", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "93ce105a00ae2126dc3a9ce4d8cbf4aeb9e60e03"} {"id": "train_046639", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["data", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "32dfac31fed7b1d63a1c0a52ba6df465e10d74db"} {"id": "train_046640", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "b269a5c3efc95e97ccb22a1d81bfc44ba426da52"} {"id": "train_046641", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "c4be7395442f5bf168048ec30bcfafb2efc24d3a"} {"id": "train_046642", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "14dc1b180d4e7eb3cb7af540066b7fbd36efa916"} {"id": "train_046643", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["research", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "c1969ba3826379e5a9465077d854fd490e3a68dd"} {"id": "train_046644", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["governance", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "d897de32c624b5562624aeae205ace3061c31e5a"} {"id": "train_046645", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["security", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "7fac73e6c3cf048fe122c26a5e9eb91d99c83348"} {"id": "train_046646", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["governance", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "d65819849d3210fc8bf449365027fdb209df9c72"} {"id": "train_046647", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["agent", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "20dbe6cc042b49a466a7eaf0fbde5f0256d59f26"} {"id": "train_046648", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["data", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "828a8448b4041e2f4fc7532213f90d5f9e18c585"} {"id": "train_046649", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "c547ff3494a701f38f9cb9bd3e75db9e7cb09230"} {"id": "train_046650", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["governance", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "415a167c56d41e9044a27dd68135e7eca1b1d82d"} {"id": "train_046651", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["research", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "c3065e440fc7d39e641f6f7c0bf6a5ad1171669f"} {"id": "train_046652", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "5ff28252d0c9e5eac8561e6ce8f4f2109eb0833c"} {"id": "train_046653", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "c89774f32a51dd2b23b397dd6cdc021557a0e5ff"} {"id": "train_046654", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 9,\n \"confidence\": 0.9,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["governance", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "a467cdc11ae6e848012cee42259737841ac00d08"} {"id": "train_046655", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "d8b629654cd6c431eb0eeb3df1383d7b28abf73f"} {"id": "train_046656", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n cost = 0\n for _ in range(budget):\n p = plan(issue)\n patch = edit(issue, p)\n ok, _ = test(patch)\n cost += 1\n if ok:\n return patch, cost\n raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["data", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "982312b2f1458721f9b911671d4c8d57638a807a"} {"id": "train_046657", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["security", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "b88b44301a1b645a58c342a8ca14fcf69a519148"} {"id": "train_046658", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "8a7af8253fa1c1e2b7fc38affa7037f70498d4dc"} {"id": "train_046659", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "dc4b01e3610ec2f653d917afdcccc9e5d4d97bbf"} {"id": "train_046660", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "65146d3da97a6d39360dca7bc6365c16dbdbf39c"} {"id": "train_046661", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["research", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "b796a8ce8f90c9f21ac9eeff3e2a25c93f8a6798"} {"id": "train_046662", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["performance", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "5f60bd1f9b8ded5303eec07d0230d2372510b375"} {"id": "train_046663", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "SQL", "moe_experts": ["security", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "9fa9a4a13b2ad883f0aade2204e91f6c8bf89480"} {"id": "train_046664", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["research", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "8db7bc704460b9b6c4fabb5a97e55200d578d8c3"} {"id": "train_046665", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["governance", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "a944918a4e626fda7032218a01c0d87b0a722d5f"} {"id": "train_046666", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["security", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "5a3ea042f4206641c6aa564adcde6b228186ccc9"} {"id": "train_046667", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["coding", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "d769c1e13198c3173c6e2c4d8e5d218382faaf9c"} {"id": "train_046668", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "2d30c286c88f4b5b13baa772a6c7e0441f6ec147"} {"id": "train_046669", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "626fdab58f2e27525f312243ff24f91c2eebf3bc"} {"id": "train_046670", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "7c1ae225a8685b7da6afab0410f1916af24dfa15"} {"id": "train_046671", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "e53100faa005b1bab41feb3e7baee78b98cf2aab"} {"id": "train_046672", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "48a5c18055426289343a5ecf90c7174a42f88554"} {"id": "train_046673", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["governance", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "e09169b21a81bd641f47799edbd8c75d7adde409"} {"id": "train_046674", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["governance", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "1f696c707565b0a657b9426458284dcf62c62910"} {"id": "train_046675", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["performance", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "d2f774e8c3d69caf9003e44e3be0fc9d11bd91bd"} {"id": "train_046676", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["agent", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "fec9a8eebee287a4bea1b24dac5c8bf9b310c770"} {"id": "train_046677", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "SQL", "moe_experts": ["research", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "010e32298e9e0fa6025197cc31f8167e309ebfe9"} {"id": "train_046678", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["data", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "c1c05ce98280cada21015e02c96f605100365c63"} {"id": "train_046679", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "329ad1c1390d3593a57d747fccf708b48fb751d8"} {"id": "train_046680", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "093f621efd7d3b75d573db5d6857a9cdd4bd2f0c"} {"id": "train_046681", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "df8991d850f9093c9b00323c892d02c050d33673"} {"id": "train_046682", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "f542dafbc6e501449a94c81d16688d840a175bf4"} {"id": "train_046683", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["data", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "3dbb1a5a48c46d36cecb72b0aaf789e4181719d9"} {"id": "train_046684", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["performance", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "989c78beb117222b2ea03ea26375a015a91f66c3"} {"id": "train_046685", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "facb817ac0b9a94f6638e579982b175d37a4d54b"} {"id": "train_046686", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["performance", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "20f656c3d51fa71d36b1c1078c1728917572d20a"} {"id": "train_046687", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "87fafe2f333eb050ebf1f8c622ee7d108054fe67"} {"id": "train_046688", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["security", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "c6e62217185c011d5668efd84767d17412869528"} {"id": "train_046689", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 9,\n \"confidence\": 0.9,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["performance", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "d92cb5963fba48e8f5c5788b4504c5799868293d"} {"id": "train_046690", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "c14c4d27b7db197aae041ff26d6e7531a2c6ba38"} {"id": "train_046691", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["research", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "364005fa62dc2035cc8d62a783aa96280a1ef3d1"} {"id": "train_046692", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["evaluation", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "d7981bd6915be2a969f636cd75d9ec5459a1d7c1"} {"id": "train_046693", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["coding", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "13dc7f5a36e567cb4ac9b2c83f0063c3864b667d"} {"id": "train_046694", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["security", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "dc58a4ab2f599e294e653b2533490269391e295b"} {"id": "train_046695", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["agent", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "9f04cf1b92638645dd958abe236c342cb4a76b1b"} {"id": "train_046696", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["data", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "060cc9c98899bf6a3f4975958b2858c82be27316"} {"id": "train_046697", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["security", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "e519097dca3e6dfc5038dbb8b0394bff9a15f3d9"} {"id": "train_046698", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["governance", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "99d8072048a9c79017532826ac196545c2f80fc0"} {"id": "train_046699", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["data", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "f2a38448954b3eab613a2fd5d4d9fd24481912fe"} {"id": "train_046700", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["performance", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "7dbb125206d44852a14bcdbd4a719747ff639311"} {"id": "train_046701", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["governance", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "105dc8fd392145a4ed4b3ca52823ab90f8049a4d"} {"id": "train_046702", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "SQL", "moe_experts": ["agent", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "370d407439060fd327994afbfcbe6b77bca140ab"} {"id": "train_046703", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["governance", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "f5bf96c0c3ca8e13edff559bdea06566f4d5dc4c"} {"id": "train_046704", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "2772d223a63a7da4a1a4cbde6efa7b8aeb5374c3"} {"id": "train_046705", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["agent", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "14100f36a452d1249ecb95760e2614388d3df374"} {"id": "train_046706", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["agent", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "2af7ef7549e1eb532dbe6d9843b700232b7fc3d4"} {"id": "train_046707", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "92b3068096725b829c5e027c59b589826ced7b50"} {"id": "train_046708", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["agent", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "6479f1d1ce203a42d5695cd41bac2170c518cafb"} {"id": "train_046709", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "3658d430cc0b167c7a15780769eb8dedba6b9776"} {"id": "train_046710", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Java", "moe_experts": ["agent", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "9723c3144157673bf701d9f1cc4a5a57942e903f"} {"id": "train_046711", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["data", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "ccb15ce18a52d18ab41427e253fe00c681051743"} {"id": "train_046712", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["coding", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "cfdca7ef97f9448d91d981ce321c714feff74cc5"} {"id": "train_046713", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "b709d796aa5a5e039b51a790adbf38178fcc484d"} {"id": "train_046714", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["data", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "ca46a3ae1a620460e94d42e8015eede3e93e26ee"} {"id": "train_046715", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n cost = 0\n for _ in range(budget):\n p = plan(issue)\n patch = edit(issue, p)\n ok, _ = test(patch)\n cost += 1\n if ok:\n return patch, cost\n raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["governance", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "616cdf203fcf7883fd62dc3cea1ca7a9c7936194"} {"id": "train_046716", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["coding", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "7dfb39af7160f9c264a5b96ff6a6a53d09304546"} {"id": "train_046717", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["evaluation", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "32f0fc5bcde60b2f625f4e34bab8392fd80951da"} {"id": "train_046718", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "768b11cdba16487b75ab86f95f6cf2b0e03e726f"} {"id": "train_046719", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["governance", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "6c5a81c0b4409e1fd77355d1345e96ef2b374b94"} {"id": "train_046720", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 9,\n \"confidence\": 0.9,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["security", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "3bf0c363b419916883b846ca1c8a05e6ea7796e3"} {"id": "train_046721", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "095e142f3e3fbc06b2eedab7bfdcafbb0b3d1068"} {"id": "train_046722", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["performance", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "080a1350ad95ce284931f076cb0d860300090459"} {"id": "train_046723", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["data", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "14c39db7a3cd0b7990ff10f4e5c7c3ebf5dcb4f2"} {"id": "train_046724", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "a2fe3cfea76edeaded90edb9768366d9e3783a5c"} {"id": "train_046725", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["agent", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "fe7927178d428fb3ae33dccbab48b1b6fab582ba"} {"id": "train_046726", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["security", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "de60a364d6e6dc7def2621e8d1729b3d846fe430"} {"id": "train_046727", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["coding", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "68c12d986d3da18fdff5620e06714f7bcd908c6a"} {"id": "train_046728", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "696a42b6a65014eae4e43cc03e908999f8454262"} {"id": "train_046729", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["research", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "007279fecc6e34a178d689520abe18503ec99681"} {"id": "train_046730", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "2c40bf61ee35834cc8c94a399da7390d2b8cc904"} {"id": "train_046731", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "fe759de682f08b5508e4fd77bad0b3f87db26fa1"} {"id": "train_046732", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "e8dd848ed1e8418219c2f88b810875912e3d219e"} {"id": "train_046733", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["research", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "673b4854214eefa6df902dc11d07f30fb9c3d097"} {"id": "train_046734", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "d36a1a9b6002de444050a66a5e1679210d0d15a2"} {"id": "train_046735", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["coding", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "098151aea76a39b602b4e33c8b6112b4a97a54c4"} {"id": "train_046736", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "2ec48ad22ad7e488a093ce5f2f9ad64ac5a6bc01"} {"id": "train_046737", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "011d09822b75785bcb14d1bf799398769439e6e6"} {"id": "train_046738", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "5a933a8831506f9f7e8aa756b369cde57e368915"} {"id": "train_046739", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "66a20271c6df4b02927bf0a07ed10d3bfcb74f77"} {"id": "train_046740", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "f2e240182a3644b57479a4c38264bf657b29c63b"} {"id": "train_046741", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Bash", "moe_experts": ["governance", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "cedd08ab8e848e8f011e0eb3f0e6ae4efea09548"} {"id": "train_046742", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 9,\n \"confidence\": 0.9,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["governance", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "5ca19eff78e256e58ab630717b1e61e99a847630"} {"id": "train_046743", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "9a561e4412cd8548cee9394248e1a86fec2e451a"} {"id": "train_046744", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["coding", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "7f27f0b3bbfd9fbf16d7c397d7abfb7f5f648590"} {"id": "train_046745", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["agent", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "ff483c90c3528ec331ad55aa05d377b06546edd4"} {"id": "train_046746", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "503e1bd4a0d7e09e5c4335caa857d042ad7f77ed"} {"id": "train_046747", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["evaluation", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "235a796ed628f47db4684d117e3f0115e916cbd3"} {"id": "train_046748", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "49c73d1159195c419099e02352fcb20bfcabec12"} {"id": "train_046749", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "e619f1a65303b4900e364b5f09182dc2c1eb9ee7"} {"id": "train_046750", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["research", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "3e14f5a83e1feb31f568176f20658ce58a1256a8"} {"id": "train_046751", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "bee16a827e510a74905ddac3b43ced3ddd6b7b70"} {"id": "train_046752", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "eea2295f86c5c8b8beafbc32d16d02c6fe0bf1f7"} {"id": "train_046753", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["governance", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "58ed28ad7065f821c4a0999b3522ce284d9b7535"} {"id": "train_046754", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Go", "moe_experts": ["agent", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "bf97512b3f3bed21bf3a1094dcb266459c362067"} {"id": "train_046755", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["evaluation", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "cef7ddbec494a7d15d052ec3f263949729e6b422"} {"id": "train_046756", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "c5c696b108f53ab792b44dde3d94aa0e77383609"} {"id": "train_046757", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["evaluation", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "765c7195ec1942e351b22631ce937551bec35ffe"} {"id": "train_046758", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["research", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "ca46a3ae1a620460e94d42e8015eede3e93e26ee"} {"id": "train_046759", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["performance", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "6db9f80ff868a7ab498724ea6b3bdeb95bc01693"} {"id": "train_046760", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["agent", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "44807b8d43f441a84590a3056a6b465426d93913"} {"id": "train_046761", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "e47c43583d4705ca8ab296763983f6b04701fa6e"} {"id": "train_046762", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "4cb541c7b729a1c5225fa0266215fca0f0c69958"} {"id": "train_046763", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "75066fa7f3e25e093f5a2f9a50699701b2d995b9"} {"id": "train_046764", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "9b2415b173755d0e30b9398f5b9657f236b77e97"} {"id": "train_046765", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Java", "moe_experts": ["performance", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "5c18895e4d4357d1e68b2ba21bf4544f74311de8"} {"id": "train_046766", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "48613139ed9ba9f4d89947f04eae0d9439597d24"} {"id": "train_046767", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["security", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "22e2a52c51b86a5c9ca5eb16f1941291544a7c6c"} {"id": "train_046768", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["evaluation", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "e0100b1a6c246288a06a170f332ba469d1e985ca"} {"id": "train_046769", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "3410a6e04a48de378e107f25355692b2431f8d79"} {"id": "train_046770", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "ad6d0e67e44e578fbfc674b59611bcbd02db4f76"} {"id": "train_046771", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "78e9d77714834547ce008824c258e7e1f7c9ec54"} {"id": "train_046772", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "f0cf42fb4d85485f3d858eb345f0a04617e6641d"} {"id": "train_046773", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "35e5287366e0fc13f0a29ee2126e37a423546676"} {"id": "train_046774", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["governance", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "6312fc1ba0f8da01cd71f874307948d09e892066"} {"id": "train_046775", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 9,\n \"confidence\": 0.9,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["research", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "d1496542ea103523142cfbea0c798015750b0418"} {"id": "train_046776", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["research", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "4deb3fff97a9b0a9f05b987056ccbf694ff7947e"} {"id": "train_046777", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["data", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "8e685b9d266bdcb0979885463eeac14a49ac29a7"} {"id": "train_046778", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "5aa110d3a8564e1a6e014748df0c7e8a8574ca09"} {"id": "train_046779", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "aad0d9309338ac3f7b1d3751a4a5e02af64321bf"} {"id": "train_046780", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "89b8564c1450c771eaafdfde65c00e36ca3eb5ca"} {"id": "train_046781", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "97d9dca1088e97a7dcc7714de3f3048a0300f0d5"} {"id": "train_046782", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["agent", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "05a2950bae1ee0126a35f406d2914325e60a5727"} {"id": "train_046783", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "e2bc21b61367e87e24c64df0ce80b8ab165ff2b3"} {"id": "train_046784", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "24557b2032c315cac5c2ecd9fcf475920d57e7d7"} {"id": "train_046785", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["research", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "4e5ca3afa2c32eff074d2bf8e6a748f9d9b7f31b"} {"id": "train_046786", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["research", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "8e8c15786b04d4b864e506efa99b8ea3379c818f"} {"id": "train_046787", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["governance", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "3d50d0388a748d1555bacc72b22cd1dc21988858"} {"id": "train_046788", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["performance", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "2ae91c85f768f0a080e68b9d3a40ab35b0da7fff"} {"id": "train_046789", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "91f7e07bda08c7f203534afbe4658f584bf80bee"} {"id": "train_046790", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "2c0f77d45d37a8450a191e3e3d0e43be43536a3d"} {"id": "train_046791", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["agent", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "05c6ee49da93b2175ebe995fa135d43bb1a84fc8"} {"id": "train_046792", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Java", "moe_experts": ["performance", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "90919dd52e99a2a2b91703ffe382fdc5985fce21"} {"id": "train_046793", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["evaluation", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "b7ae6eae2be9937254732093fe0f8ffcc953b8ab"} {"id": "train_046794", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["security", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "ff71870a52f2f390028f20822dc63f75ecc0a652"} {"id": "train_046795", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Java", "moe_experts": ["security", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "22cf7f7f6645479f3ce675b67dbbf0672a562621"} {"id": "train_046796", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Python", "moe_experts": ["agent", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "f8060fe1571de1392e1d8ed4a058b8fbec34bb8c"} {"id": "train_046797", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["data", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "c307f8cd58211a9cbf16888790c436dad45c9692"} {"id": "train_046798", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n cost = 0\n for _ in range(budget):\n p = plan(issue)\n patch = edit(issue, p)\n ok, _ = test(patch)\n cost += 1\n if ok:\n return patch, cost\n raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["security", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "e6a90ff22ecffa4a0e30e31714ec2b2317e4308e"} {"id": "train_046799", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["research", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "d0615cecb840db3dfec6e0ee0b41bdccaed422ae"} {"id": "train_046800", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "8d3034c3b0ebcaa6b5f3e5a124b4057d26668813"} {"id": "train_046801", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["agent", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "dc1e2ab7a65b8bd2cb5d8742e756f0260cb41179"} {"id": "train_046802", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["security", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "f0d45a0d7c221c0dc2ae281b11b7416a4430db79"} {"id": "train_046803", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "1e4b7b3851f1d06a5095148cb174faa62f35177c"} {"id": "train_046804", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["evaluation", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "4e0e1fd5cf9419f6d2abdf5efd1a707f9a92bc2d"} {"id": "train_046805", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Python", "moe_experts": ["agent", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "fb37479fe07da3d3a37da55892adefc122b8e8d5"} {"id": "train_046806", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "5a84e258797742bb2401ffccb7dcbb50862998b4"} {"id": "train_046807", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "f2cc09ad50fe6beb391651e9c5344a53ed33fa39"} {"id": "train_046808", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "851833a4894c07e329fad289ed3b3568dc200e47"} {"id": "train_046809", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["coding", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "a87627fe4cef7c3fbb8bc06ea05db0691e2ace1b"} {"id": "train_046810", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["research", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "55fe2915d3c8dae25aa8ac059c681d5519fddccf"} {"id": "train_046811", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["data", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "40eaaa8124dce1b16e1eb7a1674e373c7445dee5"} {"id": "train_046812", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "25d3678a2eb725c2d1460164b6d399078e5551dc"} {"id": "train_046813", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["security", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "de110e7e2eee84f6636833594343e3c03568d4cf"} {"id": "train_046814", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "cfa7cb050811f756a4f1e2b20931a71ea9729935"} {"id": "train_046815", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["research", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "505762040219f1659b725a342e8ce863879da074"} {"id": "train_046816", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "e78376fcb9271790c240616f99e9b7593b52978f"} {"id": "train_046817", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "52aa015f1bc1ac486c093967c0501553a448bb9b"} {"id": "train_046818", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["governance", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "0c1539f3e737e953717967070ef46fd21b6be3df"} {"id": "train_046819", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "9076202dbb276537389cc5a038c447862ed4b819"} {"id": "train_046820", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["agent", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "a37ad27321b6b1351200670ae663522bddb33609"} {"id": "train_046821", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["governance", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "d3e0cc5df4f3b62c1b4f8a0ff8a7d5779fd98ed9"} {"id": "train_046822", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["agent", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "1b1dc26e6603f665e7b436a8bfdc11b2a00542d3"} {"id": "train_046823", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["security", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "3d6ed9289b17a7b76b0c049327043070eb031881"} {"id": "train_046824", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "5d44460d9d49f71980a710ddcc75d928019e88b1"} {"id": "train_046825", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["governance", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "e43cdc8c99e8adffdc42dfa59c71bf45a7a92182"} {"id": "train_046826", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "c9bb2df5d2cf979c2de4d98aaacc1d67f76f1b61"} {"id": "train_046827", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "782e34eba2f30cc8096770f8ef32d4185df768c2"} {"id": "train_046828", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "431281a1ede60d5777b71e7704ff66f856640ff2"} {"id": "train_046829", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "4586e20b5bcee667d0f78854b361b66d1d4dde15"} {"id": "train_046830", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "f74a406c59a1bc28bd9f2a2dc7d9e35e60c43e0f"} {"id": "train_046831", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["data", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "6757dca802d00a09a0beeb361d38ce582f51d91e"} {"id": "train_046832", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "6c2161ec4c3ac885ec0cee65a0167f8fac1b5ae6"} {"id": "train_046833", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "cc5a7f3339b10c2a430ee5ffc12b502470d6f91c"} {"id": "train_046834", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "2c0c1dd586e30a526d1792dc6029576ce47027b0"} {"id": "train_046835", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Python", "moe_experts": ["coding", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "24f20415325ca18c2901a592b51250df02e13e37"} {"id": "train_046836", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "44383ba935ee3a956a3970445d190b2e6609d831"} {"id": "train_046837", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "f2d0fdef5e8dc1b22983a56343fe59d006ee6459"} {"id": "train_046838", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["performance", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "6018b4a569874fdab152796f02e727efeb4ead2b"} {"id": "train_046839", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "94c2f16af9c2e75629f4e4ccf75613b7b36516f3"} {"id": "train_046840", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["research", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "10a5262879c1957864999aa9776cd8253b520366"} {"id": "train_046841", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "318e80a88228500294769543a21053cc294e1490"} {"id": "train_046842", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "3fcd2f08e097fd36735bb9ff6ca676331eb4e135"} {"id": "train_046843", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["research", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "839b2b211adc2c1c9088a047882ac7b9739e13ab"} {"id": "train_046844", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["security", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "73782d816df047fcf07c759903ae97d45007e949"} {"id": "train_046845", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n cost = 0\n for _ in range(budget):\n p = plan(issue)\n patch = edit(issue, p)\n ok, _ = test(patch)\n cost += 1\n if ok:\n return patch, cost\n raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["performance", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "86edd581571f9c1cd4898db47fc1ea802d150a36"} {"id": "train_046846", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["governance", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "1bc9ae1d3cd2e53bafd9091cb32a8e3b6976286a"} {"id": "train_046847", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "643839cde00ebdcdbff3fd28e55c52931562d6f0"} {"id": "train_046848", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 9,\n \"confidence\": 0.9,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["research", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "f920a24c5219f4c6efe97b226f63434bd300c0eb"} {"id": "train_046849", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["coding", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "5aed8cea893eac75d0add5c2dd3b7e0b802db8a8"} {"id": "train_046850", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["data", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "f728b449eaf2056c2c29bd0baeb2e82bbcdc33dc"} {"id": "train_046851", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["research", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "30b5d97926c8b786510b91489fbec8706fe47fd5"} {"id": "train_046852", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "2fccb0c5deab1b293e3a067f96e32a212dff9f31"} {"id": "train_046853", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "b8f0b61878034ee1b54ccc319350ca3dc48d957b"} {"id": "train_046854", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "dc5d9d9a0e89a7f3e7891928275ef1d3cd212502"} {"id": "train_046855", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "d55e80f32d296207512270ca852d689877e6f0ae"} {"id": "train_046856", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["security", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "a24b3f285cda550dfaec313f86181c003147e3a3"} {"id": "train_046857", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["security", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "8bcb5b04540ad72d9e9a1200738587cd7eb05525"} {"id": "train_046858", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["security", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "7e16e57757cbb0a6bebb49d4ffe2e2abedd9d684"} {"id": "train_046859", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["security", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "adc4b6f7836422b53e9808fde6c6b8bd2f9aee80"} {"id": "train_046860", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "eb962871ec8a1baaa9b8de5641fabd222a34ecb3"} {"id": "train_046861", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["coding", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "72602667dd04f1b0c8ecd6e2fec4cd01907aad36"} {"id": "train_046862", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "c315ec07bcac88b043c2bbe28bdcab93995cd3c6"} {"id": "train_046863", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "ab852b2db3058cf21b48d51be264e74c06e2bc88"} {"id": "train_046864", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["governance", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "89f40a2039dbd083b7d39195b9cbcdf26fdd6a78"} {"id": "train_046865", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["research", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "4b72c5555142ff3f520447f3949f83171ffe2f7f"} {"id": "train_046866", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "df4c60723157e4394dc0ed8f605a5db4d49b949b"} {"id": "train_046867", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "318e8b8ea6c125b532235481460ebd92296a2108"} {"id": "train_046868", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Python", "moe_experts": ["evaluation", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "a42b138983a384ea4ef9327b4b687732eb2b9440"} {"id": "train_046869", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "C#", "moe_experts": ["coding", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "cc24798f77ff93deffbf1cd6658435e7fad2b16d"} {"id": "train_046870", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["security", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "d25515a77dc6791bed56993c644965c8b70c56fe"} {"id": "train_046871", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Java", "moe_experts": ["performance", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "688bd9c277ebe8c1506ca5d8c3c035c8d09fc55c"} {"id": "train_046872", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Java", "moe_experts": ["evaluation", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "df6b45fc4b3020d3638168294ccc0038a0769669"} {"id": "train_046873", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["governance", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "954631db8d44ac7938d77fb9d8248c804ef478eb"} {"id": "train_046874", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["coding", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "2a601b60b48730d05dce86c835faec9975849f07"} {"id": "train_046875", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["evaluation", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "05ac77711c79499ae41fb177f1d018a13563fa42"} {"id": "train_046876", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["security", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "c2313b35e965fee0070c06cb3ede4c89639c95a7"} {"id": "train_046877", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["security", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "3e14721dcfd915628a81f972b399eb8da79524ab"} {"id": "train_046878", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["agent", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "6935b0455f285c02d6d35c570e769cf6be4dd713"} {"id": "train_046879", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["security", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "0ee0c79cf3f1a20d862acab8a146a93bd209d17b"} {"id": "train_046880", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "189f4e5eeb3ec37f278d4e5b27d48921805f405a"} {"id": "train_046881", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "72b192b937b2e14cc36b84e8a2863ddeb0a4c5d5"} {"id": "train_046882", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "78cbda13de06ffd08080d2b4308b5b03f030fae4"} {"id": "train_046883", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["agent", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "cd97c5dddd067bf7f014de5de57d32a3d2929770"} {"id": "train_046884", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "0a96c5ff398582ec76ec006e60779ede328174d2"} {"id": "train_046885", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "f0b65fee1a80c0d34b2ab0ed1c3d1c582fc98cfc"} {"id": "train_046886", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "52117e20acaf9870d35a68eff0c3da6c5bfa27bb"} {"id": "train_046887", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "5f602875da7bfcb4b4413b5f2887b7f3ef41aef3"} {"id": "train_046888", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "946c5268a8ca9abaef3b914694867b6cf54a16db"} {"id": "train_046889", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "99485c937a5629aff90d6ced9fa9848c76ddcb47"} {"id": "train_046890", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "fbeb962dc28310f9aa7b094322cb839ba86ec312"} {"id": "train_046891", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Python", "moe_experts": ["data", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "d23e3dfc7c65df4168e941a145a87e4afe9484f2"} {"id": "train_046892", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "2a3a05c6bec3e0562d38632a82ced67dde72fc42"} {"id": "train_046893", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["performance", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "6bf28e4314b461a5144bbb1fd4dd9b35584b2a0c"} {"id": "train_046894", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["evaluation", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "9f1778a91431c3da438d5c438a6889755c696947"} {"id": "train_046895", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["coding", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "6378971f88e520cf65fa6dca3ee12f9bca8c533a"} {"id": "train_046896", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["governance", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "f7b6aabf7ea65262445f8ac7417f90945e37f6da"} {"id": "train_046897", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["data", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "365aa600457c94ab9242705fb005a7608f63a751"} {"id": "train_046898", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["evaluation", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "494534723e51f2a463cfde7827cd1d040e94611d"} {"id": "train_046899", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["research", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "84adbc9698733f8e64606f341f13cab7b62aa631"} {"id": "train_046900", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "626000982917b37db06770a6b38eec18f8705e56"} {"id": "train_046901", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "SQL", "moe_experts": ["research", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "195beffcd97e1b57aa813b2c5766ccec0abbf522"} {"id": "train_046902", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "5e24b084a772abdb4142e93b877786ad953f79a1"} {"id": "train_046903", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "2bd66369e420db151d0ff41493dfeb0e82cc7e03"} {"id": "train_046904", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "0247b493cdaf649323b959428413189fc9f36f2f"} {"id": "train_046905", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["security", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "d7bc156642229da6d4cb400a6fe582bc435431d2"} {"id": "train_046906", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "05d49af8e0efab2fe06fd62611f226cfceb4596b"} {"id": "train_046907", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["coding", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "39fbea79ae05e4d690654faa4b64b1deac9ebb93"} {"id": "train_046908", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "d225b742e01776cd9e97b49b4a51f42e0061f357"} {"id": "train_046909", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Java", "moe_experts": ["performance", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "76deee9e4283ed68910950c5b5b186101c591d18"} {"id": "train_046910", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "add960a85c0b3b2de909503cf30ab05faf83f507"} {"id": "train_046911", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["performance", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "3a040a15f2d024244458de24a7ae966a0eb451d9"} {"id": "train_046912", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "404f0bb8aeba66a014c3060fef83dcd550a33b23"} {"id": "train_046913", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["data", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "c7490d039ce580a5391d3f40519ab1a9d595d138"} {"id": "train_046914", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["research", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "144aef08461a7497efd5e3ca42cf5969b59e921b"} {"id": "train_046915", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "4944c4ce4acf2c9cd176d36d3cea519e56d72994"} {"id": "train_046916", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["coding", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "42921215f4517a0c361ec97b9cc9296493c9c08a"} {"id": "train_046917", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["agent", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "3ec0a502539b14b6f8d2bbba1496cc1c1885edcb"} {"id": "train_046918", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["agent", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "37b25790a88488fa9d9373bd608a6a24fa58cec7"} {"id": "train_046919", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "cb725a3f9203a7ec9766e512fe6a239da55e5100"} {"id": "train_046920", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["research", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "5587280b315ba3a53d220f95a4ce7da21d9143fb"} {"id": "train_046921", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "6771fdda578a9f042280f344125c35efa7885edb"} {"id": "train_046922", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "621746dcd2e9608b3af2fef9a02f08747781070c"} {"id": "train_046923", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "5931682a68f7fd7aa52195beb71109b35ae0bea7"} {"id": "train_046924", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "79957774504228ac2de51be58f58cd864ca42aee"} {"id": "train_046925", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["governance", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "68f4967c9b152b262cf70acc50ec599c9f768214"} {"id": "train_046926", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "01c3ce1e07faa4fb9c3d1603ec4edc2f06371b6f"} {"id": "train_046927", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 9,\n \"confidence\": 0.9,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["data", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "d2ab3959fe74e787c1c38e055c6613f4df964e7e"} {"id": "train_046928", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["evaluation", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "a998bc8b814cffc83331a370f4e3dc3ecab7ccb2"} {"id": "train_046929", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "SQL", "moe_experts": ["security", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "9f2b84ecc0e1c66fbc94dbc1c17f3f8e3d5bc4d9"} {"id": "train_046930", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Java", "moe_experts": ["data", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "efdb96a0e3d22bf887f6584081adc4b5b9e1fc6d"} {"id": "train_046931", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["security", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "6a3c280e74bfb1f4c8a14dde4cceca5b6eb8e920"} {"id": "train_046932", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["security", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "97221ca3e157f277e0db42864bc1fd0069250fec"} {"id": "train_046933", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 9,\n \"confidence\": 0.9,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["security", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "3f92c619430eb179a284791a5f5141b2ffdc732a"} {"id": "train_046934", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "0a6840d4ddad75e0d364d52255cde21f60850a72"} {"id": "train_046935", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["performance", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "7c47594b69982e70516ccae274b887d3bc5c6c19"} {"id": "train_046936", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["coding", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "dea6f93a71c3ab79357fab396e4629b94ab2df38"} {"id": "train_046937", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["governance", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "23ecaae1f12ce138c7436b99c8d17c5a16617099"} {"id": "train_046938", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["governance", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "4ee8eb10bafcc22cfff35be36f70c50354df7024"} {"id": "train_046939", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["data", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "5b5329c39c24831d5a04c981a97da5d65cff8509"} {"id": "train_046940", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "62f1b6f4a460bf4a742d7a839f9a69c1cd3937fc"} {"id": "train_046941", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["security", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "e38f99467f241c05b590ecd672d4a31632cb4fa1"} {"id": "train_046942", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "30cd7d205e415dd74e83740414b7ba45c74a1648"} {"id": "train_046943", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["data", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "96691c1a1b8d7877351fff22dad61cc61ab836ec"} {"id": "train_046944", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "67581b9c2389174c49ce5d5f96510d705289aa34"} {"id": "train_046945", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "7fac73e6c3cf048fe122c26a5e9eb91d99c83348"} {"id": "train_046946", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["coding", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "853969ca7fa1a28e2b1b53cdf8928db57d174161"} {"id": "train_046947", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["coding", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "024082c106f535de9e0157a852cc9f5b76165499"} {"id": "train_046948", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["security", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "cc24267b8ea39ee7b54891b764cd13c9fcf57480"} {"id": "train_046949", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 9,\n \"confidence\": 0.9,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "99ee4af581e848070ea264f3dbde16d862ee409d"} {"id": "train_046950", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["performance", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "a4904a44b1a1cc1f1c4c3f2c7cd3ed6c6bee9bc8"} {"id": "train_046951", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "SQL", "moe_experts": ["data", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "01f5f92542021de524e0a9413c3a1af8c9cb9070"} {"id": "train_046952", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "d47514ebec3fbc5eb01b247bf3744b3e4244e621"} {"id": "train_046953", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["data", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "38f1a0abe1ac4fd08c77a2127de90867e225a4c0"} {"id": "train_046954", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "b8abcefa71b25ce959b45e7a669a6ae46c8b033e"} {"id": "train_046955", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["agent", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "c8d94c96b8b55711f8d0fc248c66edcebeb9cdf5"} {"id": "train_046956", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["research", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "fe19ca7af37f3017f12ac87226cad68e1fd046ce"} {"id": "train_046957", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Java", "moe_experts": ["coding", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "9fd55bfefd2fd8482807c69bd7a706ccea5b05dc"} {"id": "train_046958", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["evaluation", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "d979bbcc4cf286e6922a27f17c374996d31f7fba"} {"id": "train_046959", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["data", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "93f1b679f5a2ebe7e0e88698728bb6309b98e845"} {"id": "train_046960", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Java", "moe_experts": ["governance", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "e8c59c7e5e824de0d9ea6255f971776f9e8c3c0c"} {"id": "train_046961", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "9d6cf952a2c5f52adf99e07350933c8c8813e4b3"} {"id": "train_046962", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["data", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "a344abd8592e10d753bf9298ed4d94ebc80f010c"} {"id": "train_046963", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "2ef3748285da8790cd5709f6c6e0ccb4745f1349"} {"id": "train_046964", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["evaluation", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "1e95a7ad23bdbd53b53122276acd98165beeb676"} {"id": "train_046965", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["research", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "abb3c459c96a01c069ba1e7d566a9d5dca81c547"} {"id": "train_046966", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "55ac8e868de881c5f4e444008b58ca2a2c053aa8"} {"id": "train_046967", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["performance", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "ca568902b61ffcb0f1e46aceb37605072487ece2"} {"id": "train_046968", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["research", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "e927d7a23de86170a7b23b38114e04ab50d8a988"} {"id": "train_046969", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["evaluation", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "b13152c6a4e1e1ee03bf9c6064832245afefd900"} {"id": "train_046970", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["research", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "de707f6b2a70fb4dd3ebcd1d79481fea98aaa01d"} {"id": "train_046971", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["security", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "489c9d5b976fbdf069f0a8af6ce2b466bcb71fad"} {"id": "train_046972", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["security", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "f07b9d4f5ba14b3f09ca8e7b7acc3fca7be40362"} {"id": "train_046973", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["research", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "f3072a6ea5ceac8fdbea33149e9a59de3c973d07"} {"id": "train_046974", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Java", "moe_experts": ["security", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "56c774ac28bba89ae1f6f4aeda6dd255b5219a09"} {"id": "train_046975", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["agent", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "7b214346998baf0b7373de99197731dc0ad371f0"} {"id": "train_046976", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["agent", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "4118d4634888e7f5a22ea5305a834cbee8985074"} {"id": "train_046977", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "33d86473efd596a6218b68e41bcb70c48a3f3a72"} {"id": "train_046978", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "acf87143a5c9b01f57feca51ec4cee6eaece064f"} {"id": "train_046979", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "fc354ebfe8f0d514f9a3eb5fca7b8105f2e13666"} {"id": "train_046980", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["data", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "c6975f758b727964e0e8c132a2e97e9b250c30d3"} {"id": "train_046981", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["security", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "6cdcaf1d868c0a13a11981dfb57bc25c19d78fc3"} {"id": "train_046982", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["data", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "4fe7fbdd82bdcacf696ef6a66ba205d4764a80ae"} {"id": "train_046983", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["research", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "8d5951b09108e1fc7ce9dbaaf1d9d9aef3556fad"} {"id": "train_046984", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "3db46754738bde6cb8d655ccc406c256b30ce0ec"} {"id": "train_046985", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["governance", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "d4d9667f78d6b4954b3fdb17a7dca9f454c78fce"} {"id": "train_046986", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["data", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "4328c2175db85abd4fdee1c2d36fac97bda52a50"} {"id": "train_046987", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Go", "moe_experts": ["data", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "790c643c70a0de7682bdafb4b01f3a8222771d79"} {"id": "train_046988", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["research", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "a9343ad9d5e6464a387a95e39b6a82b0d31cae0e"} {"id": "train_046989", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["performance", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "3b8b7ddeec7032e7aaf592da871ce1d29ea1b037"} {"id": "train_046990", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "0e41438e2fdc8cce8dff5bad1f66a0c7b8aa6056"} {"id": "train_046991", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "d28074892ab4fb82a3bbedcf99ea2ed9b148abdd"} {"id": "train_046992", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Python", "moe_experts": ["research", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "fb816796a39b0f57e8bb60a780a3c7a4d4752777"} {"id": "train_046993", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "368c27bdb40c4577f4b3ca72aec9f297294b6550"} {"id": "train_046994", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["agent", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "b028fa0455aca61428201fbb08df98fd5e82e70c"} {"id": "train_046995", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["performance", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "d1ed87dc1fe5687f378932a827b4aca26386c152"} {"id": "train_046996", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["governance", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "c0dc60caf7d16084a84cbd1f3e62d9504af2faf6"} {"id": "train_046997", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["evaluation", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "978e8299753f655638ab0e465225e827cf895137"} {"id": "train_046998", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["agent", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "4c8cda7d22abce6c24f28a478738bb9f786fd046"} {"id": "train_046999", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "29af3c423f9e4fb39e47b5fa2619f1415ef177f7"} {"id": "train_047000", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "c2d7c1cdd70d0d18df38f47e2a11d91eb918ad8a"} {"id": "train_047001", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "061fcfe72aee4f01f845f0ab731895b0253846b2"} {"id": "train_047002", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["research", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "1f6f884455f078afc2723ed74b978bbf0945215b"} {"id": "train_047003", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["agent", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "565c9efeaf45f5b73bf391bb8562d52b569bbdea"} {"id": "train_047004", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "52db5b2ebc68a39c76a072e6b7c91ed63b6e1400"} {"id": "train_047005", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["agent", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "a3c7ab82662e2b5f9d1bfe20755e545454ab9214"} {"id": "train_047006", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "244b32b108d5f9bb0e7baa73c8e83989a64b81d8"} {"id": "train_047007", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "C#", "moe_experts": ["security", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "11500a44c573016cdedf9cdbdc2e4df5036d28ec"} {"id": "train_047008", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["security", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "d7e02ccda21533a39e6b05d932de08220a96c123"} {"id": "train_047009", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["research", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "22362dee6d78fa1f4c2d43491442825631245187"} {"id": "train_047010", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "b0ef32bcb1656983d2fb0ffa14a5c516029e2042"} {"id": "train_047011", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["research", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "325b2b5f13c3644e338c077a775b0fd10034e47a"} {"id": "train_047012", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["agent", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "4bcbf40c8af6797089dd81f4df4bafe2382020b1"} {"id": "train_047013", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "346be1b2fee0cb7045c0965d9265325f42f5620a"} {"id": "train_047014", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "3e409f039b3fb158be4413da98292b3536ef5b10"} {"id": "train_047015", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "22c657ee649b24b27be29ca99b70c62f8a13ba83"} {"id": "train_047016", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "e59680cf1f12e89264047abdcb79bc7de45d050d"} {"id": "train_047017", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["research", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "bfae200f5a7acf4418a80d6164745a008cb9ca9c"} {"id": "train_047018", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "85fece1b48f9d4f8334d106d1939474d47e4bc5b"} {"id": "train_047019", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["agent", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "c3de58d72de4ee2b9f6d41faa0a62808076d1ae9"} {"id": "train_047020", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "3b5f18064cc4427d3bd3f60168681782717e9d87"} {"id": "train_047021", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["data", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "a5a1327cbae1387934643ae308805164f1a5459e"} {"id": "train_047022", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "11d9ad9117023d72cfad1c68d62c13d48f6dc5a5"} {"id": "train_047023", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["research", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "8fe10d93c943c5b6711a1aa9c27c4d2597cdc205"} {"id": "train_047024", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["performance", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "71fe1017b1aecc3a58bc266758a89b90286fafba"} {"id": "train_047025", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["governance", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "acf16c77362bc1a65addf854bcd613f61bf3c2c8"} {"id": "train_047026", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["coding", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "7c65efebc8f02b49847e44aec0fe3ad854c04744"} {"id": "train_047027", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "acc12ec69f7f6e28b6056bcf1f94f7bf2d649a70"} {"id": "train_047028", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "e7ef816766a3e9f1a117ee85c1343ee306539b8a"} {"id": "train_047029", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "c970982f979f3086122a23d78cd6703a7cdd71d4"} {"id": "train_047030", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "424565d7e8db39562040afb125af8f2731eff37e"} {"id": "train_047031", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "23a0b07a29204c9edb32ba077dc239f355ad8243"} {"id": "train_047032", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["governance", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "220a45166b53f686f83bedaa300ffb970d7b4e79"} {"id": "train_047033", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "a4c8be7ca28490dbecd982a21bf7fb6ae4617fd9"} {"id": "train_047034", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "3f7df3140f55ec53fc7f2e59e767e60e9e6b70f4"} {"id": "train_047035", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["data", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "4c7ff7a8dfd0787332f0480aacca9cec93b69d30"} {"id": "train_047036", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "0ea63b7dd5d06d69ba839fb8817e688f13a0d8a9"} {"id": "train_047037", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "5e568295c03f17cefe5845316120b69868990146"} {"id": "train_047038", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["agent", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "fbe02c19deb701725bb8b3060e8dd6353654d2e2"} {"id": "train_047039", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "58c1fa532ea768a661833cfa541744572bce6e70"} {"id": "train_047040", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "ff0d96bb8b3fead610cb00b6cad2b0e28c95396a"} {"id": "train_047041", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "6cdc4cf34f3b9e7c403ee590e7f78f106ad11a5c"} {"id": "train_047042", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "35d8aac1cdb2981c6728a99cec0ad9f5556848df"} {"id": "train_047043", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["agent", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "e302b1fe904a059294e575177aec11fadc3d65ea"} {"id": "train_047044", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "2bcbd4d69d306720e17139f82efac1997dfbf8e7"} {"id": "train_047045", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "f4659f623d54b663116467a2b429a9006c3852fb"} {"id": "train_047046", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["evaluation", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "b8f52dae1fe0ac735508ab0409c475f041ccc41c"} {"id": "train_047047", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["research", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "1e52cca2a078a6e7dbf1631f384ec782a2dd9742"} {"id": "train_047048", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["agent", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "e55e6c1d77db1a73d342fee5f76a96f11f114fee"} {"id": "train_047049", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n cost = 0\n for _ in range(budget):\n p = plan(issue)\n patch = edit(issue, p)\n ok, _ = test(patch)\n cost += 1\n if ok:\n return patch, cost\n raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["data", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "9e7f1853b98c5000e5cc82f1dc82300351f048d4"} {"id": "train_047050", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "2c62bba27be26cf8101269b40651b4397970c092"} {"id": "train_047051", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["security", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "d8eacf250c09828948a0548021945ee359d90c66"} {"id": "train_047052", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["agent", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "e3b334cf6c1af2d521037215923c5e160aeba598"} {"id": "train_047053", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "bdf542d60e96ede5a18cc024b2ae1dd43cc86041"} {"id": "train_047054", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["security", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "00922c0644610d2ef71a152c29572c1695a16a88"} {"id": "train_047055", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["governance", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "2da4fc709ca7001eb874c74f449bdd0a56b07bcb"} {"id": "train_047056", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "661f74f3642eba5f273e079ec0bd467372adff76"} {"id": "train_047057", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["performance", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "5c75cf9474888fb61bb189beb14f555afb1c4137"} {"id": "train_047058", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "3bf6fa62347af14ce14b7d44bb043928768f766e"} {"id": "train_047059", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["security", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "335ea61cb14c7f2b7523dd0033acdc8ce7cc2201"} {"id": "train_047060", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "57655d196b85ab47c0827617bdf66ed191ecd900"} {"id": "train_047061", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "c8db01edbf25496ad99a842357e7e56c46ebc5d3"} {"id": "train_047062", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["security", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "0ddc60fc61864d3962bcdec4e8c88e263fc8adc1"} {"id": "train_047063", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "29428ea1b1ff782f3c3de06a8d6ebe51eb06d1a8"} {"id": "train_047064", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["agent", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "9f7b300e83e75f76b591bb5535c1f44fa33a35f3"} {"id": "train_047065", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "a1154cc33b0a592621aa53b3410743206c0aa35c"} {"id": "train_047066", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["coding", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "a6e4b6e9ca2a53a0e7cb1b4734347e10b13652e2"} {"id": "train_047067", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["performance", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "9788c26098279184d8e586dc28024de2d50ffe52"} {"id": "train_047068", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["performance", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "d0141eed2eaf673e80969a2c6e233419af851107"} {"id": "train_047069", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["security", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "635623c0a53be7c9eb60887761aac1691db21f4d"} {"id": "train_047070", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["agent", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "b063954e6763b016d1c3deabf83493c592449b39"} {"id": "train_047071", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["coding", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "122e0cc1d8cd6fad6d34b75d5af713eeb5bdb45c"} {"id": "train_047072", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "616abb073df88ae0d57a9735d64cac98ae7558fd"} {"id": "train_047073", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "07278b24fac8e7f18a9a18d602dace725bb06906"} {"id": "train_047074", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["agent", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "f1afaf618ed8eed0c2a79c385d5ec1860c18ad4a"} {"id": "train_047075", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "61c6b062f4cc89b7175eeb7bcb6922e2fb421563"} {"id": "train_047076", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "f7a9c579b586c56f1fbe640c0d8835a5164f31b5"} {"id": "train_047077", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["data", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "58c1fa532ea768a661833cfa541744572bce6e70"} {"id": "train_047078", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "97863b59fc0907220679af6c71e69e3db9ffc4db"} {"id": "train_047079", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["security", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "c5066c32a774d009605cf605dc241a9abdf95520"} {"id": "train_047080", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["security", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "313a79c02ab96e81b338bd132d22ad4e3818dc7d"} {"id": "train_047081", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Python", "moe_experts": ["data", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "b25439f94471459472a48851975e7b0a9887362e"} {"id": "train_047082", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "d167ef22f2f65ec5b387700cc64276d1a266542e"} {"id": "train_047083", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["coding", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "5aae1556d7b46f19f98f126bf26918f7df9281bf"} {"id": "train_047084", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "bdfdd2bcb8b3ce897c0a5a3aec2f42ae0480aa30"} {"id": "train_047085", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "3ad69ff46f138e988fc66c329808c12277d95cc8"} {"id": "train_047086", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["security", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "33a9bc1160df62e000f9577a05fdeaf19be57f42"} {"id": "train_047087", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "C#", "moe_experts": ["governance", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "36d7d03301a206b589492f4068a845b10043306a"} {"id": "train_047088", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "e368e8fd10e3b180abfdc2e17e490f1096aa8533"} {"id": "train_047089", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "65fe7380be8622cc6c720f42ec901d387e7b2f5a"} {"id": "train_047090", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["coding", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "2a7d7e2e097419834d7697c4f61049a24feaeeba"} {"id": "train_047091", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "75d1b3a34f8f5e9d0623473ecea1df0376f25dd0"} {"id": "train_047092", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 9,\n \"confidence\": 0.9,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "dae2362e9b7ced19336befcf7922f3b2b6ac3b35"} {"id": "train_047093", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "9d7196f91776b4a14da930f7d5073feab8eceb8d"} {"id": "train_047094", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "cdde60891cd6aa708572f6d1cb1320e6244086ac"} {"id": "train_047095", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "21fa29a3e9846aeadd1146f53062713cbb053520"} {"id": "train_047096", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["coding", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "117f413bce7725bddda8ab90a5fcdeef615c3122"} {"id": "train_047097", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["research", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "74eb982984f08378ac4fd79713bfabcf736604b4"} {"id": "train_047098", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "6d0e306ea444320f399f7a609bc64b6f698936e1"} {"id": "train_047099", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["governance", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "760981963b09cfe791bff811799f942a1b508718"} {"id": "train_047100", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["data", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "2c52efcb8d0be30e5499aa35762448f5a0c9b3f8"} {"id": "train_047101", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "43fed9166b8ffabfceb049547c7b0687da114fae"} {"id": "train_047102", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["coding", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "3b98f9e012d8cdb72e9689203084434452009114"} {"id": "train_047103", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "74bc710d868338c41b2d6e45ead2d11c444b5e76"} {"id": "train_047104", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "0f7e72a07ff727af8b145aa3d05f9adea569e5d6"} {"id": "train_047105", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Go", "moe_experts": ["performance", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "2a0a8dc4c9da431dfc9c07c1c3bfb1505e56176c"} {"id": "train_047106", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "d7f29584f71097a335b5a9465c4fd89157a5fcf5"} {"id": "train_047107", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "9f2c2dc8ea5d6b1b9219965565da4d85463a3ff4"} {"id": "train_047108", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "edd44bf9b144f85cc16e19fe8f5eaa62e022f10d"} {"id": "train_047109", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["performance", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "d2a89726f7ac317902cdb8b811a1f5741e7bf12c"} {"id": "train_047110", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "576f543c39fc1d49a2ec1cc8c4ee227cd1af73c9"} {"id": "train_047111", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "00f807037862536f5c271a3c1a619d67ef1d36eb"} {"id": "train_047112", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["security", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "b2bdbddc168c68111f7b831da5d9526df5b3b7aa"} {"id": "train_047113", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["research", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "6dafaba564acc884f450757c78069b3d1f398afc"} {"id": "train_047114", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "C#", "moe_experts": ["performance", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "e2483a66dcb7f1059b662f2d065f2ebeea042930"} {"id": "train_047115", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["governance", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "515d3301fb27dd4a0571dd1511ca70a270c08ce9"} {"id": "train_047116", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["research", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "f755910e3515df5e8193acdebdb8b04da1f3e48d"} {"id": "train_047117", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["security", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "64e4199db274c5d918fac3b092e39541b99099a7"} {"id": "train_047118", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Python", "moe_experts": ["data", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "28f6daf27785eabd944f0be6d96e1516b2ed9dfd"} {"id": "train_047119", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "5b505e2b78cca8fa4bced8c07fc87f8e78c5260e"} {"id": "train_047120", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "93ce105a00ae2126dc3a9ce4d8cbf4aeb9e60e03"} {"id": "train_047121", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 9,\n \"confidence\": 0.9,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["data", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "ff9941e88b2b70816910a6214d776a913f8a2a37"} {"id": "train_047122", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "cec172dfa3dc12f0136923e5bac01b15030830af"} {"id": "train_047123", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "b5750e210b022853b7d845417c3bbfd2e9de616c"} {"id": "train_047124", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Python", "moe_experts": ["data", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "9f0677bcb554a76ce8bd8c03421b653036a6f7c9"} {"id": "train_047125", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["evaluation", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "bbedb87efaaad718f73a4b234fa79165b0d4c422"} {"id": "train_047126", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "d013cd06fb3e9b59b70033a09686b140759ad07e"} {"id": "train_047127", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "95733a8785a91f0ddf19303a47b95aafbeb3c59a"} {"id": "train_047128", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["data", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "f795d2596016f39474d1ed323c62cc0a496a72ec"} {"id": "train_047129", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "e7ea49b24165988af212cb486aeb7dc6f1d5e635"} {"id": "train_047130", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Python", "moe_experts": ["performance", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "714aa81e4a8f0b9739e3914c96086a52a331dc09"} {"id": "train_047131", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["performance", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "7d13043aa55abfc83f4ad14f2e70f3deab074321"} {"id": "train_047132", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "4cba24632bd28a15ecdec8e7e1d3f0c52c1c5bf1"} {"id": "train_047133", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "997e6018fda10394eed43c6466d5b684e120e308"} {"id": "train_047134", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["evaluation", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "04eb598cdeba1986a4127e9c6961de4761ea6929"} {"id": "train_047135", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["security", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "e2287f24b0a36825d98987ac0c17209b521ae711"} {"id": "train_047136", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "ba2e3c573dfebf0574ab200f5900100d979fd3e8"} {"id": "train_047137", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 9,\n \"confidence\": 0.9,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["agent", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "093e2d4a225446f95dcc0f5f2602b07ab5f74700"} {"id": "train_047138", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["security", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "db2b90b72f15cfbac610539843f90f99772aa40d"} {"id": "train_047139", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "4386d72f48fe50b72d10e5af82ec663f54bb21cc"} {"id": "train_047140", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "e5025ef17300c15c80c07da3a07e769c4b1314c6"} {"id": "train_047141", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "b0018520fde97b5d3c91ea6b05bc4b868aa78811"} {"id": "train_047142", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["performance", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "09e366834423600bcc06e29004d2299a3b5d0d2f"} {"id": "train_047143", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "0571f2d09047e0cb113dc127b14dcdbfebe5827a"} {"id": "train_047144", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["research", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "f8c0af0ba3a7691ac4d35c8e11b63313224a2be3"} {"id": "train_047145", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "895d8f33cb6f06508e6f7c1928200dcbe8cece5d"} {"id": "train_047146", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "b2924b715989dcea6af5d0240d3e786089e6f6b2"} {"id": "train_047147", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["security", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "f082631a5bc6b544a1e6c8b530ba528b8f198018"} {"id": "train_047148", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["security", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "20f4f08036fb94fe1f71e70709db468dc380bb85"} {"id": "train_047149", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["research", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "6f8d679774ebd4f04ca2a9e1a03dca56b3fbffeb"} {"id": "train_047150", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "41da8c97baffc49d4bc7032e1a8178110a586475"} {"id": "train_047151", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "dd952bf7fbcdc98b1bcae62ffd653e4610e28126"} {"id": "train_047152", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "4ca0cc6de3ebf4539eb27a412362917276744479"} {"id": "train_047153", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["agent", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "82fc4cddd61ed918102b893d49c6fd23394bcc33"} {"id": "train_047154", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "32da3bc8442c39d261a892a5e88bc4aa3fa08792"} {"id": "train_047155", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Java", "moe_experts": ["agent", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "4ed3ddd0013387166535b151a7b8df552b064d50"} {"id": "train_047156", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["coding", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "388614f4961b121c93fa17eacf13775ddef98351"} {"id": "train_047157", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["governance", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "8c784d393c1af6619ec32b2c02561c82a1a245c5"} {"id": "train_047158", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "2faa058dc6677ad1c4db9e7ca6bdc410b6eae348"} {"id": "train_047159", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "9ec3b7835d76d1e361a5f00b5940eb2d8d692bbd"} {"id": "train_047160", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["security", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "3a14f4bf37a10ef2988d485364192a13b47e93a8"} {"id": "train_047161", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "78f6430397b75ece3a78d079e73e995d08986122"} {"id": "train_047162", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "76321b76dd05e242a073b7e257e45437b5ebdb0b"} {"id": "train_047163", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["evaluation", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "4e8ef6268f5db6ea21be08295ca968624dadd7f3"} {"id": "train_047164", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["coding", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "d8b2e05e5d3b240e4867dd90e44bb6afee4d49fe"} {"id": "train_047165", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "9cf62812d6b4945192ea65a7e5d099864b978572"} {"id": "train_047166", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["research", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "f907d9fa2399b051137a7123f4002dfb9ab24407"} {"id": "train_047167", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "990a916e10d33c61a2734c7dffbfda93c98655b2"} {"id": "train_047168", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["data", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "7c9730a91441d7b5eb4fb4b76dc992988a6fa276"} {"id": "train_047169", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["data", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "099083244e40cc204a92642a709b826d79e4a19b"} {"id": "train_047170", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "f97c19a032c2febe9b583e8ffc418312035d77bc"} {"id": "train_047171", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["research", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "ac7c3d6e4301621b39aa0bcf0a76931b6b430dba"} {"id": "train_047172", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "580e3764f58b2edfeed84d9399a5a142739289b5"} {"id": "train_047173", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "565447c84b27498be01e4d6ede89cd67f258aac1"} {"id": "train_047174", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "cd00b9a27e67455f45f5ba813678807cd1f7b214"} {"id": "train_047175", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "a0bba63280841063c47f433dff0f42271ee9beb6"} {"id": "train_047176", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["security", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "a2b04111458c333ad8c1bd8c3921819a93f36495"} {"id": "train_047177", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Python", "moe_experts": ["research", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "5bd2e353c84cff45036a63f0ceea65532b820fa5"} {"id": "train_047178", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "d1515646625631d4e7f4c82bb3caf09b937f6330"} {"id": "train_047179", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "c93fe9ebfe4b9c699b0c9e415a217710263e51dc"} {"id": "train_047180", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["performance", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "74b0233e1463b22f865f7dbc479e19f0c3a99efd"} {"id": "train_047181", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["performance", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "21af030df0e9ca15557bbb105d3eee5e6d97cd72"} {"id": "train_047182", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "c1015ff527e942a0f3535026cacac43bb6ac1942"} {"id": "train_047183", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["governance", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "0d844f389887bb25a1055edc10a0244b805695f0"} {"id": "train_047184", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "f66848762706a1b97d359d92410257cd6cb0a701"} {"id": "train_047185", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["research", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "8bc3c851b85b971056e28ea9ca9336b73890fd20"} {"id": "train_047186", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["security", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "0b0322fc8ccb919922117b56f26ba832afb4e8fc"} {"id": "train_047187", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["agent", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "9bc653128daa38a548dd5ff34e1bd05f242615ea"} {"id": "train_047188", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["governance", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "5cb7500060afeef4120c40c2e700caa9c7ce4bb8"} {"id": "train_047189", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "540b290d5d9fc8ce317d17905f2e4db11d90eb38"} {"id": "train_047190", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Java", "moe_experts": ["coding", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "1505bc9c477b6436533d778be7ea8a28e68f0a1f"} {"id": "train_047191", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "0a9c58f07d8a66075d7a758c19900ea91a04b63d"} {"id": "train_047192", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Python", "moe_experts": ["data", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "ded33c4506768d04a757e79d95af1431ad217a72"} {"id": "train_047193", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "fe86ec164457fa111239d49e79b54eeab81d46e9"} {"id": "train_047194", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["coding", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "db53c45ce0c05cc0a9fddf6eeff822c7aa67da3e"} {"id": "train_047195", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "bdb8767fe0bd8a9a205a0ba88398b8911a404d1d"} {"id": "train_047196", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["security", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "ec4d20419fe820dd1aa38c1e7fe1238df6cc52f4"} {"id": "train_047197", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["coding", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "c06bc6792bb352d2eace33865ebf6dd0ed2ad0e2"} {"id": "train_047198", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["coding", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "72b3f4e1b2485c525f3f2e1f2c2899326eea8fbf"} {"id": "train_047199", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n cost = 0\n for _ in range(budget):\n p = plan(issue)\n patch = edit(issue, p)\n ok, _ = test(patch)\n cost += 1\n if ok:\n return patch, cost\n raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["research", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "fb9ca4d5b9fe8574655ea304f0263fca5af65c7c"} {"id": "train_047200", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Java", "moe_experts": ["agent", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "d6a8b7d8886de3be6cf1746ce2216e390773b54e"} {"id": "train_047201", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["research", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "e8d64fcb3b1386bc150595bab4ca791d3339891c"} {"id": "train_047202", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["governance", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "de6f9ca24521bbf3cac1b0a7fd8aa43ecbfd8e3c"} {"id": "train_047203", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "8afb61fb764949273440b06bab8c114eb190c68c"} {"id": "train_047204", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["governance", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "c8e69f511f3840c5ba35ffcc1039bd20a12895b1"} {"id": "train_047205", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "63cfbcc751193d1d7e04dc6b7bc351de9cfe62f8"} {"id": "train_047206", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "3c90900baa8aa9cdf5736020c2b35cff8c77e879"} {"id": "train_047207", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["security", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "c543ee711198640b815a3f924ce56dd728c2c924"} {"id": "train_047208", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["research", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "6a934f6e7a23f8c1695128d560564bebef6cb35f"} {"id": "train_047209", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "aa81f086b7e6342c33e9c5c053ec13b7f7207633"} {"id": "train_047210", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 9,\n \"confidence\": 0.9,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["governance", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "43710723122fd183914bc365852b727c308a367b"} {"id": "train_047211", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 9,\n \"confidence\": 0.9,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["research", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "67e18be16bea57f312e064d518d285fb458e435c"} {"id": "train_047212", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Rust", "moe_experts": ["coding", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "e44cc4a3bf0f8595a0535804bbc95b0b5d42097d"} {"id": "train_047213", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["agent", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "3eaa74e10f3c20ecaf3578968757e953dcabd520"} {"id": "train_047214", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["evaluation", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "7785ea2e8277b140ddc68bb39de684c2bfed3a99"} {"id": "train_047215", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["research", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "3ff91086e7ffd0d929635e6a53201e2446832cb2"} {"id": "train_047216", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["agent", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "816ecf468d299bf6fb2876d055ed465b66809f1a"} {"id": "train_047217", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "5094eb5ea2372f382d06f787f2ecc8bb7de8cc53"} {"id": "train_047218", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["security", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "bd067d3d796724e8ba600dc0ef0929d49564703f"} {"id": "train_047219", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "e314cb0002660b0c79891ed7f5e2e93127ed8e07"} {"id": "train_047220", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["performance", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "2bcb4caef884a0c35ff842bcc92d71fba52093b6"} {"id": "train_047221", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["coding", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "c665a38a032b7512e714761899d71f36a9fea460"} {"id": "train_047222", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "C#", "moe_experts": ["evaluation", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "af4ea35e355aa9f030030bf170832e82c192c038"} {"id": "train_047223", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "6add7d9277d50b4836a76d398264545546c599a9"} {"id": "train_047224", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["security", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "4eadef5f2880ae840c7b01910d39d17bec3e551f"} {"id": "train_047225", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "C#", "moe_experts": ["agent", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "be198fbbd5051a98696a46e1e9879dfa096dacdc"} {"id": "train_047226", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "ae6d4a8fab7023c0fce0a97f8471f7ff6c3a24c6"} {"id": "train_047227", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "a09d0e19d25a6737c7566f74ef32d0cdac723b9f"} {"id": "train_047228", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "fb2e7132bb0e59090c5558108d29ca479a290313"} {"id": "train_047229", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["governance", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "62005d4a4288b8a1f34aa429ccd984308df6abfe"} {"id": "train_047230", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "50e0416b62dcab61944c9b00d6ac2d08d20d7735"} {"id": "train_047231", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "SQL", "moe_experts": ["research", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "c67c143e3e0eda7281d4888ef85b2f2b31159b83"} {"id": "train_047232", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["research", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "d355335e83eadf5177952573e5a56bf81f7eecf2"} {"id": "train_047233", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "b16b2599ac3901d62b50c44b94bc174878a60f99"} {"id": "train_047234", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "d8cc09d93716a91c0df3d025cf41ac48357519bf"} {"id": "train_047235", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "c734d54771f242aadb44d99a2532604059b97cbc"} {"id": "train_047236", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Go", "moe_experts": ["governance", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "bf6878ce3dacf271288abd3314159f0f5096d204"} {"id": "train_047237", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "916f8895aa83c5ebcfed036411ddc8af76ccfe20"} {"id": "train_047238", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "bd25d1bb807db30e2fde4c8c2e33a98906310155"} {"id": "train_047239", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["coding", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "db419701eb9b8819898d52379cbab52e782d18ab"} {"id": "train_047240", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["data", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "5636e5f934a8ae06a9a3ab3340071a76021fbed6"} {"id": "train_047241", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["security", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "e893d91bdc2b43595991c84e86bb4611d7724fd4"} {"id": "train_047242", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["performance", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "599cb1e559d013052c94a0a22cb44511ed6c68cf"} {"id": "train_047243", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "3d466372c67a8210216d71ee576239c925ddc60c"} {"id": "train_047244", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "4bfc0f02fb72feefbab9a87d5079257e99822d47"} {"id": "train_047245", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "SQL", "moe_experts": ["performance", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "85361f59c8d351e71b4d137feda2b07648261732"} {"id": "train_047246", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["governance", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "c4812370e6878b16cf994f0933cea6944c0b8d77"} {"id": "train_047247", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "fd2c7185d81b46f941d0e3ba8e3e774a7bbf6077"} {"id": "train_047248", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "004521fbd948fd3d409def0341d807db25a6e275"} {"id": "train_047249", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["data", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "28b6c4aab232933797a96e08995b2e0e6f175766"} {"id": "train_047250", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["research", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "4cbac7af2fa2ab551611b317ccec7e93b3032ed5"} {"id": "train_047251", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["security", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "7f76fa257c52243d0662a3d6dabe3c52b2308b92"} {"id": "train_047252", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["data", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "33330e1315bd450b9757849902cc365375bcdbdf"} {"id": "train_047253", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "f8e2f5caf55c92f933f97214db504e478d7bd5fa"} {"id": "train_047254", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["research", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "df97888e042c97f257bdf3a4c935d0d5cb326036"} {"id": "train_047255", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Python", "moe_experts": ["governance", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "a0c4127d0297ca3b96134d75cae15ceaee26d647"} {"id": "train_047256", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["security", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "f6059ad05a94956716d90cecb4cb2732a2d6a097"} {"id": "train_047257", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "bfd9d51c689651a4d914a8fa6a224ea1e9be83ff"} {"id": "train_047258", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "362c506e4877a07cb298d5f4b34aca9043e51d7d"} {"id": "train_047259", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["research", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "4f327429bfe8b46dfaf74f2de632424f9aaef496"} {"id": "train_047260", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["governance", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "15a6795b30315b87a65244226def1cfb4e41e8f2"} {"id": "train_047261", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n cost = 0\n for _ in range(budget):\n p = plan(issue)\n patch = edit(issue, p)\n ok, _ = test(patch)\n cost += 1\n if ok:\n return patch, cost\n raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["performance", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "2eb9374edf23caf55ea17fa59b657c32f581e503"} {"id": "train_047262", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["governance", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "7bc7cd908215dac7b522a7c00a8b4312086c0558"} {"id": "train_047263", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "751d5b9e2965825ac565726e25709baa9df614d6"} {"id": "train_047264", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "3b88eabaa75baca13cb178768fc73b7a4b7cdb98"} {"id": "train_047265", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["coding", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "57774dc15c1d820859e7e272371871957e158eba"} {"id": "train_047266", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "f5575fc6818ac86d66168760f255454f5f163fcc"} {"id": "train_047267", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["agent", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "9f04cf1b92638645dd958abe236c342cb4a76b1b"} {"id": "train_047268", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "d6261ae344b31591e7f2b8c6410068239e3c2cfa"} {"id": "train_047269", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["security", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "32c7ac9b07a94afdbe4de7eb3596da0dd7de7c8f"} {"id": "train_047270", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Java", "moe_experts": ["data", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "f54423c4c817f84b87c6f79e8df12f96b472513f"} {"id": "train_047271", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "cbc5cb52e54f9c0edb50c8035b47dc9c1a4ce039"} {"id": "train_047272", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "6419db58ed6329fb88c0f561bf5cf1a408f1f98b"} {"id": "train_047273", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "534b4df7a648e1162342e164439c5288f53224d4"} {"id": "train_047274", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["performance", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "a925d59d5c78870d1ebaef4f25a4b1d4ea68d51a"} {"id": "train_047275", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Rust", "moe_experts": ["data", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "afad463200311fddfb61eafe0765266e68e1056a"} {"id": "train_047276", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["agent", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "443b5b1f9abdae35552c215ac78a892090f00be8"} {"id": "train_047277", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["performance", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "74bc04c800824daa1e19ed81b2014a9593a37107"} {"id": "train_047278", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["security", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "21b57bae0dcdcc3a6062669eeccf49e3c933f07b"} {"id": "train_047279", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "f2a38448954b3eab613a2fd5d4d9fd24481912fe"} {"id": "train_047280", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["governance", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "1913bf65a9661766550e92d8174b2e24a692642d"} {"id": "train_047281", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "79679a2565d38789e0e7452078cb4e8ec34458a5"} {"id": "train_047282", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n cost = 0\n for _ in range(budget):\n p = plan(issue)\n patch = edit(issue, p)\n ok, _ = test(patch)\n cost += 1\n if ok:\n return patch, cost\n raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["research", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "4607d6d5357135e047fc12666effd1514887df97"} {"id": "train_047283", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "cf6192f6660cec01a253d2a3c60e13c428a78b70"} {"id": "train_047284", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "51b61e698ca85b8d2190c33cdca2b1363ee9f153"} {"id": "train_047285", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Python", "moe_experts": ["data", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "ec62b689497d2e048b0b000d80bee98130d78d32"} {"id": "train_047286", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Python", "moe_experts": ["agent", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "b1491183be337c3552b630d53c4fb7b4e197b88b"} {"id": "train_047287", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "5d2d3bd14e5e24f89e46a0763f302010ec96ace6"} {"id": "train_047288", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["coding", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "4360d5f5020345bd9bfeb8442c4623d4c849ccca"} {"id": "train_047289", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["research", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "621faea85ff1691009f2b655a82371939da630cb"} {"id": "train_047290", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "27bd0ee398566ce6a624d13a7127aa230c727a8c"} {"id": "train_047291", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "16357bc1627084dce4f4151af9fbfebc0c68156f"} {"id": "train_047292", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "bed1ffa2ad04866c8e1780982f2875d01e7fedb0"} {"id": "train_047293", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["security", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "5a694566098a938644b6abb967d25fda71463248"} {"id": "train_047294", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Bash", "moe_experts": ["coding", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "5d5595d0e6b1edf0b6cacab9d534d876a72b8513"} {"id": "train_047295", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "257eec780851a5c1e33ca31d012413bb7974f7ec"} {"id": "train_047296", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["data", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "a8d93b2186051e90e1712e9954f7d583809f6c0c"} {"id": "train_047297", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["coding", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "c393414786e250ff6da1ce4b400d73c73354c83b"} {"id": "train_047298", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "85c0617b2bf0afe0b0a9db0b4ef31fb20e851494"} {"id": "train_047299", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "681eb3fae676401c84592460bbe48034b84f6291"} {"id": "train_047300", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["governance", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "a0ea08be45e61915669f9865d5799986cf9d829a"} {"id": "train_047301", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "7efde6ccbd972356e4231ea118f5c5ea36d924ba"} {"id": "train_047302", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "0b6f3abae0eaf564928e362c5df95c3842e5ff5d"} {"id": "train_047303", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "77771b1609aaad4d4687bc9de23ef108ab469ad1"} {"id": "train_047304", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "3535ebac7603a77cd21f5147f5dbe3a34c850106"} {"id": "train_047305", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["security", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "fbc24fb5b3d47da2f67ab6bc641076dd6ffc5707"} {"id": "train_047306", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["security", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "5b40e7fc8ae61ea1a565a612662deb5cd9ab204c"} {"id": "train_047307", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["research", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "4efc958436354093bb6fb6f3e970960ebc9f5f11"} {"id": "train_047308", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "11f3c74e0c57a8bb075906c7c6ad33efc4feafc9"} {"id": "train_047309", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["governance", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "e27dda8422c74bd772a791a9bc55581e52e038c5"} {"id": "train_047310", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["governance", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "bde16e8efefa1988e3b1c83546d1a17d80f74aa9"} {"id": "train_047311", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["security", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "3d610dc5cefeebe5e974d809a3339facd0459221"} {"id": "train_047312", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["research", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "1c143ea2f2ba5ee668525e564a18d7b07e96ef9e"} {"id": "train_047313", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "a79db43995cfe1d10e381fdce56640b34634abc9"} {"id": "train_047314", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "ebed0e76bcbe04ec7a839261549df06181cc0f19"} {"id": "train_047315", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 9,\n \"confidence\": 0.9,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["research", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "97f206f9e9b3ca8946df475b741db79c829832d6"} {"id": "train_047316", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "b4f98b0c54a6b3935d3ddadd05666a4a42fb6ac2"} {"id": "train_047317", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["performance", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "23d625d82faf9d55335da0b0a2fb8d4a6c4361b9"} {"id": "train_047318", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "C#", "moe_experts": ["agent", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "757e6bd621b77b55e1ae9b683def78fb8b904514"} {"id": "train_047319", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "a321ff685d63fbaf813550c6fe0057d63b3dccc5"} {"id": "train_047320", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "1104df2e56331cb7946622976d6bc11cbce89560"} {"id": "train_047321", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["coding", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "53dacb5f2152e854f5ecfe66596cc8ab3ddda375"} {"id": "train_047322", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["data", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "d07bc29156fe0d916e3c340d33a24a6ea8cfba66"} {"id": "train_047323", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "25576dfac22b1f046fd28ada46ec052672cad002"} {"id": "train_047324", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "e606cca094644f49ca7767542f30303147a62c69"} {"id": "train_047325", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["security", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "e38e5da78eb195642defd1aa79bd1794031d59d1"} {"id": "train_047326", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Go", "moe_experts": ["performance", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "ed528c9bce17f3b408b4552b246d2efd5a112770"} {"id": "train_047327", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["research", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "9913852c96e08c292095181b215c0eac41d38a96"} {"id": "train_047328", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n cost = 0\n for _ in range(budget):\n p = plan(issue)\n patch = edit(issue, p)\n ok, _ = test(patch)\n cost += 1\n if ok:\n return patch, cost\n raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["governance", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "3161097cc70b04bef9407dd343a71551e1dfaad3"} {"id": "train_047329", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["performance", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "efcada577410cf93b7ad13b422a8146091afda56"} {"id": "train_047330", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["security", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "2b1b1ef1c8aa61e1e76fecce860ae29434427975"} {"id": "train_047331", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["performance", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "15949b46c1972a4fce10c96691311c069d5b1792"} {"id": "train_047332", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["research", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "4838681b1e0def6ac9733611bb0c85085eacf89f"} {"id": "train_047333", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["security", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "9c713a249236eb9a02443e9c1e800b2d6634acf2"} {"id": "train_047334", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "1f2cdd5a9a29cd6fb2ebce87bc2ac12740654060"} {"id": "train_047335", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["evaluation", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "a7b33cb192ea9b68902d3ea32657221551303ba1"} {"id": "train_047336", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "e514092845226633efc4fe820f92c7e2f8eb2c53"} {"id": "train_047337", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "daa456c846956f86dabfa9fd2a6ead94186aa051"} {"id": "train_047338", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "689fb0b45e2dd13167b639de9967cdd5d543d77d"} {"id": "train_047339", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "2baa9ce0ce3cf93027124b3b46cc87d4a020e3b0"} {"id": "train_047340", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["data", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "f20550c8d686cae2205376f9eb82bfa80f466e0d"} {"id": "train_047341", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["security", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "9a561e4412cd8548cee9394248e1a86fec2e451a"} {"id": "train_047342", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["performance", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "caa8aa0bc02f351cda29e7ef2add939aad204011"} {"id": "train_047343", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "8d6148ef79f8bccd951b0d299f3ae6db6edbafef"} {"id": "train_047344", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n cost = 0\n for _ in range(budget):\n p = plan(issue)\n patch = edit(issue, p)\n ok, _ = test(patch)\n cost += 1\n if ok:\n return patch, cost\n raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["governance", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "35aa639c03ed1633860bafaca052a0b9b984e663"} {"id": "train_047345", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Go", "moe_experts": ["data", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "5e526437c3773e7d02cb829d299940f82948bd64"} {"id": "train_047346", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "8ec4b3093c1557af91e42e463919b69c52dbbb4d"} {"id": "train_047347", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "d3eec790d19022d14ca6b422b303aae5c2a86f15"} {"id": "train_047348", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["agent", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "0e855dbcb4546d3176e59359547f636a89601772"} {"id": "train_047349", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "cce79a4443c15f6f38b8370c3d08fc98bef6e46b"} {"id": "train_047350", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "15c5898bd899875a1e04729378a5c6eed7e2832f"} {"id": "train_047351", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "445d9edee3bb15a6ffd6f1536038ca155a0097b2"} {"id": "train_047352", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["governance", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "ef0ff2783932194e58863b59d2222eb4b3b13deb"} {"id": "train_047353", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "2ab17f50df5ee830d501d849dd9d3f6f43c3a24d"} {"id": "train_047354", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["security", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "12750ed383657e41cf9152038fa34d912d20016a"} {"id": "train_047355", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["security", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "61a73b41560fa2a27ce320254e6ce229b3d1c498"} {"id": "train_047356", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "8c3e85483192848a5012933bb9847016e078f7a4"} {"id": "train_047357", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["security", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "77c6209b4bc9e738e67f17b10c3334c1c2635266"} {"id": "train_047358", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["performance", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "0cba8a8d2707268c64f4e5aa59860d373180f3ed"} {"id": "train_047359", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["agent", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "ce8f31e33dd3771f2bd9a9830e818894bbe21845"} {"id": "train_047360", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "5d404b57c9ec5d3aca241806bda4853ceadf0577"} {"id": "train_047361", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Bash", "moe_experts": ["research", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "52bfc84d92a70824450ccc3f6883be49225d2379"} {"id": "train_047362", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["evaluation", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "e719eb1f3c3a0e62c042789b718a1c4386789a95"} {"id": "train_047363", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["security", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "6a9bfa92458d830be232f40522c46cf2b0cf6698"} {"id": "train_047364", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "41791728e1db3b6ab26e58d39eff9c9b41d8a0c9"} {"id": "train_047365", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "822b51e748e41016ca12f05d9e182553579ceed5"} {"id": "train_047366", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["data", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "2f8112607725779979219aa92d0983a96dec53f5"} {"id": "train_047367", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["security", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "1f696c707565b0a657b9426458284dcf62c62910"} {"id": "train_047368", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["security", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "7d4acccad0f1e3409762e699e8a372566e699bd1"} {"id": "train_047369", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Java", "moe_experts": ["agent", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "22d3a91fac3eaf53668c4473de391042d9d73909"} {"id": "train_047370", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["data", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "826f68390981e9644621dfa643e7d8f30d23af38"} {"id": "train_047371", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["research", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "7459607ef0c53e3187e39d075b05249f920b161c"} {"id": "train_047372", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "2ac64879e7be3bc380b6f2dad84b63146267aad7"} {"id": "train_047373", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["research", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "58ce7e391891147164f8afb45c385ef787d71a92"} {"id": "train_047374", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "85ecc5c902fef2b3d6844a172e160e810f207610"} {"id": "train_047375", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "1fb9b3e6c33aff7e2811bd074fd7e1ce5f415625"} {"id": "train_047376", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "3042a6e82ef2325163d55f2bba475f3f11d8a3c3"} {"id": "train_047377", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["data", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "353685efe6ce33f95582aaa8632f736248f0f932"} {"id": "train_047378", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Bash", "moe_experts": ["security", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "6e8573c2b2693a767b76952fb2492af1a16908d2"} {"id": "train_047379", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Java", "moe_experts": ["research", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "9f3a96a5649e09801b262ab0cd132653d072f3f6"} {"id": "train_047380", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "55be245127c25ae089d367edb076174319d7e76c"} {"id": "train_047381", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["research", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "67d0ec7ece74d0f4bf606939374ad71718902402"} {"id": "train_047382", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["security", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "dff8e18a1f1c3968f7ec0ec576b483339ad1ddcc"} {"id": "train_047383", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["data", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "21f84c2eeb18191f7d5091a9323424e26aa22bc0"} {"id": "train_047384", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["data", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "e8b5fd42c136218b59e78c0a1453d80b54f0098b"} {"id": "train_047385", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Go", "moe_experts": ["research", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "c930d1e8383e7b4f1d51501e223fafd88c7521b3"} {"id": "train_047386", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["governance", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "91c4d658a1f49f72387cc183080b57411bc708e9"} {"id": "train_047387", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "0cf1573458e6d696068ca88490e6330fc6969f70"} {"id": "train_047388", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["agent", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "7b2855e80fcc801d41114ab64cfa7469bc13fa10"} {"id": "train_047389", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Bash", "moe_experts": ["security", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "2147cfaca21582587128846374359883238998ea"} {"id": "train_047390", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["research", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "bc5dbbd347d6db856c723f234f43655407fd3280"} {"id": "train_047391", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "0caaefd06d1f113d6234fd73b8a953e09c738d78"} {"id": "train_047392", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["security", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "e1fe3cac7a9a2e14b483471e672f04b4bb45171a"} {"id": "train_047393", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "b07c397dac5f77a711e915c58430dc7d02cc42f4"} {"id": "train_047394", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "b82fcd95572e91a702f690352624fe19fc73f723"} {"id": "train_047395", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["performance", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "57e69637dd0055d576397cdb1beec08c008db4e0"} {"id": "train_047396", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["coding", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "15004a0bbe8b84fafd14149c572be858f22d6c77"} {"id": "train_047397", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "e8035d432d6fdf2c373f58217cd72dbe0da5df70"} {"id": "train_047398", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["performance", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "812865355dd7f880079da1d8a1d012d6ccfd4ee1"} {"id": "train_047399", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "b8f52dae1fe0ac735508ab0409c475f041ccc41c"} {"id": "train_047400", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n cost = 0\n for _ in range(budget):\n p = plan(issue)\n patch = edit(issue, p)\n ok, _ = test(patch)\n cost += 1\n if ok:\n return patch, cost\n raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["security", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "9e7f1853b98c5000e5cc82f1dc82300351f048d4"} {"id": "train_047401", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "cb140c9e2552b2f17d7d44ff86d4cf2e1165d563"} {"id": "train_047402", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["governance", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "a7307a35656ec442b074d10ac4979174bad9ce1a"} {"id": "train_047403", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "0058b4852fb8f79408d5ed9a86bf9c80c5899b30"} {"id": "train_047404", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "eaf7cc440439b9786cc081de38a3eb38e8904e84"} {"id": "train_047405", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "75dc639526f675ab1b257fed87fb5f4e4ed867f2"} {"id": "train_047406", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "5e70966b08f0f1f65ee946bf216b91862583ba64"} {"id": "train_047407", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "31537559f17e8a8a2e9a7021ff759915d5c02f49"} {"id": "train_047408", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "19ca688939f96083c46c248a97d085fb8a4aa716"} {"id": "train_047409", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "aadaf342210c5f5289f6f5cfece18660d1f0cf60"} {"id": "train_047410", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "3d4af73174a066046fa2f3f67a856571c7a3eec7"} {"id": "train_047411", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["security", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "dbf9f0c17a9255aad8016cf29ba6b957ddae8622"} {"id": "train_047412", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "916f8895aa83c5ebcfed036411ddc8af76ccfe20"} {"id": "train_047413", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "ca94b89cf24ff48fc7cdfe45e95d3f8468150914"} {"id": "train_047414", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["coding", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "f2ce35a24e1774c0de7863bc09fec714c7f2e51c"} {"id": "train_047415", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "c93fe9ebfe4b9c699b0c9e415a217710263e51dc"} {"id": "train_047416", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "c68881973e87a6871fee85537788111f2c6f4abe"} {"id": "train_047417", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["data", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "dab6ba8bb305fac32b23724392b557db203e9fe6"} {"id": "train_047418", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["coding", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "16d61e7497bfb849c309da0b434d7d0104f93245"} {"id": "train_047419", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["performance", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "da0cd82555e9c25bb076b84a509a5adc83095151"} {"id": "train_047420", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "f40cc848c54405bb1733129f6ed6944f7e231498"} {"id": "train_047421", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Python", "moe_experts": ["governance", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "b8d4f02bc144cce05078098aaaf4d456f9eb0784"} {"id": "train_047422", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "09e366834423600bcc06e29004d2299a3b5d0d2f"} {"id": "train_047423", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["governance", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "8ba7289e014d5ddf156a2b96d4992eeeeabd3ad0"} {"id": "train_047424", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Bash", "moe_experts": ["coding", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "abd8813f0070090ecca01b3cd6a5a0b4f15a7392"} {"id": "train_047425", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["security", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "57f0a3721e3e58d64559d4e481e2cc9f81b75abe"} {"id": "train_047426", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["security", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "486daf944454aa91a227bf8c156f73c168fa0107"} {"id": "train_047427", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["security", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "242d769ffe9268cf8f6edc01509b12f1ad0cf4c4"} {"id": "train_047428", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 9,\n \"confidence\": 0.9,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["research", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "a4257b3a02115b07930bf966fe788c037b620011"} {"id": "train_047429", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "b3d46694475e9485d77428e05e75b2b56716e5c1"} {"id": "train_047430", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["research", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "d8f37a5e60f037147c4bf6f5e35e25f10a8c1f35"} {"id": "train_047431", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Bash", "moe_experts": ["performance", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "fb19190830afcf8fadb103f35f4be7e13243d7cf"} {"id": "train_047432", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["security", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "d7f1d69f3ab2c01afcc8b61e34a30d6ec1953caa"} {"id": "train_047433", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "3b3b260f18f31901f21e35c805e72acd1c167f92"} {"id": "train_047434", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "39aa4351685d1611e1bda3739aea99ea63898d01"} {"id": "train_047435", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "560259e481d4a94e23f87f2c69f10de42e35c3a2"} {"id": "train_047436", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "8664751b738c71a395e8b2927a9cfa4b3890ba3f"} {"id": "train_047437", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "0363ef714f9a2691f614b40e09925d3778a9df6e"} {"id": "train_047438", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "f4f5a021e7042ff4690c24ecdfb6deeb3640ec41"} {"id": "train_047439", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["research", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "91276eeba7da5751a32aee0172a48a5d391bb24c"} {"id": "train_047440", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["agent", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "e80bfc26b3b81dd21092b3bbfd1a2206d5eddb07"} {"id": "train_047441", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "2789be09592473d88ab3dce2a6f7d82d05623179"} {"id": "train_047442", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 9,\n \"confidence\": 0.9,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["governance", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "e3b196f0e76d9396bc812f10e999717f1707a948"} {"id": "train_047443", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["security", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "6ab6948c109e67e9d6bfea0337141db33beb7861"} {"id": "train_047444", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["evaluation", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "f304e3fe9ea5f8c12ee7459e15c2185c9b4ef5f9"} {"id": "train_047445", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Java", "moe_experts": ["evaluation", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "52021b947a481f90a60c02c8de24c33669b5ce24"} {"id": "train_047446", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "60f5906cef2bc006e32fcf38458e8cf497ee8d01"} {"id": "train_047447", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "c1b231baaca9d6d5cb65141e600dea66810274ff"} {"id": "train_047448", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["agent", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "a09f5498f5ecdbe10d64ba7400eeb3ef2ca3bbd4"} {"id": "train_047449", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "be9e134c9ac31073aced3a7b7be99f3870c40401"} {"id": "train_047450", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["research", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "926ca0d356130117ba6395d8a02949220fa435e4"} {"id": "train_047451", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "576f543c39fc1d49a2ec1cc8c4ee227cd1af73c9"} {"id": "train_047452", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "7fa3a8fa476aa00b0fbbc91b26a9d76eccda9a63"} {"id": "train_047453", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["performance", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "c3295256f367cdd8f15ba93dea199c6cd675b34f"} {"id": "train_047454", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["security", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "79807e43b3b9351228745d905191f79aafb57c14"} {"id": "train_047455", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["evaluation", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "fa2c39be828ae73c7c5a8fced6c152b281684f89"} {"id": "train_047456", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["data", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "7e16e57757cbb0a6bebb49d4ffe2e2abedd9d684"} {"id": "train_047457", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "2809d5345ec2ec27cbbc0ce697f0f1bd7f8b98e4"} {"id": "train_047458", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["coding", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "5864d8362585a76694f4e0806584228ab8b06487"} {"id": "train_047459", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["coding", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "da346e94ab19c3aa953f96388a30d76fd5eee4d6"} {"id": "train_047460", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["security", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "d8449e1906d94574406ad8104045aa2d58b694de"} {"id": "train_047461", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["coding", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "cf6581fdb43b4ceeac86728a34c7a0e6ece695dc"} {"id": "train_047462", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "25d6353adc35801f4992e43bb3a35ff10fa51922"} {"id": "train_047463", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "02a56219a402b28f1540a0b0d22295aeef09a2fe"} {"id": "train_047464", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "a2d0565068fa1cc53d755abea691ce478c176664"} {"id": "train_047465", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["performance", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "fdfda76b521f2828ecea955eb46eed36e759a178"} {"id": "train_047466", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "46e0e54e693bf39137d83468cb28c8dacaaf23f4"} {"id": "train_047467", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["evaluation", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "3cbb2ff0c24ec5426d8922ec13c47b211b4f77a7"} {"id": "train_047468", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "a87627fe4cef7c3fbb8bc06ea05db0691e2ace1b"} {"id": "train_047469", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["agent", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "d6f2e459424d7d4335d5336ad44250bf5301c52d"} {"id": "train_047470", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "2bedfbf83787c6c05e8bee77c8e5d6fad768f1b7"} {"id": "train_047471", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["data", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "1c59f1aa0903c3024ea956340a183c49aa3f4cc1"} {"id": "train_047472", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "169e473b368190149ba9425fc4fa5097175a4dbd"} {"id": "train_047473", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["governance", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "01b16b75482c36e37468d97c217e91c6b9ea0634"} {"id": "train_047474", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["research", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "b9456ef7145462904fcbbfac722374ff82adbe85"} {"id": "train_047475", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["coding", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "f43d7aba3b42f74669ddd713ee82a1dafe93736b"} {"id": "train_047476", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "8987b403d86bb56e316e717a1ef2dfe45b7364d4"} {"id": "train_047477", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "f90fef6b16a389b1b8da2b2722a3d5f980663b86"} {"id": "train_047478", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "87785d212ff2d46fee7308e60f09ff2e2802a25f"} {"id": "train_047479", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["research", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "29bb2c67b00af5482a469be40ddf8b3944975a25"} {"id": "train_047480", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["performance", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "6425658c6e042781d2f091e6be77d8dbe6b8d263"} {"id": "train_047481", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["governance", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "725404e627728ace187e4e49295eb7f4073556a3"} {"id": "train_047482", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["security", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "8b04e992efc8477c17a8933fb39bad194aaaa928"} {"id": "train_047483", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["security", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "16fd72d33a52c9c39d502589976b830d3a803299"} {"id": "train_047484", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "0302afcbca45d4961b34a940726c0e8f52ae1472"} {"id": "train_047485", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "9dd6e4246a82398840e1f794d3312d20d9302b9e"} {"id": "train_047486", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "42dd6a8d5d9c2c50380e1cf555c9d2e2c72139c5"} {"id": "train_047487", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["security", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "117f413bce7725bddda8ab90a5fcdeef615c3122"} {"id": "train_047488", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["evaluation", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "eac7428f5cfe2ed21e6a00201d93850c355579c3"} {"id": "train_047489", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["data", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "76731075d4eb4e5b69d996eb6f85f8787280b96e"} {"id": "train_047490", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["agent", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "60d98d21d2537be3a1db96d9392baaa9419adec9"} {"id": "train_047491", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "26bc4bf44da031211d1b19d9996400186f34f1e4"} {"id": "train_047492", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["performance", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "e1b3a86276459ea7633dcbb2fff90eec12e331ff"} {"id": "train_047493", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "SQL", "moe_experts": ["performance", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "18cdcd4b48fd64129e42d1d18edf2d8a985ec520"} {"id": "train_047494", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["research", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "4cc8751ad9b167fe9a3636d544e04acdbf4a0976"} {"id": "train_047495", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "e49d4971c85cde2b4aee4d4483c8ed538457b436"} {"id": "train_047496", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["research", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "57a7cab42f6a2ac6635e6b4710fb59693c62c1de"} {"id": "train_047497", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["agent", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "38dc2be3261dc25d9b171b228ea452546fe4cd5a"} {"id": "train_047498", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 9,\n \"confidence\": 0.9,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["performance", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "76fedf100e672d216f60656cfacd06e6726b5d45"} {"id": "train_047499", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "d1d9e361a8f7700f5d27dae7d3f49eaaa4d10145"} {"id": "train_047500", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "e355ff449fe017b3a0e28112d13bc0f7e46432c2"} {"id": "train_047501", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["coding", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "f3072a6ea5ceac8fdbea33149e9a59de3c973d07"} {"id": "train_047502", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["security", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "1a95a8f5745c0e4de4a4ce97e8f38e09eaa7dfd3"} {"id": "train_047503", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "efb9403f6659faa2722fba499f74d90bfd6fe482"} {"id": "train_047504", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["agent", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "718d79fd0863bff974d5518d5fe35fc4cfef6e64"} {"id": "train_047505", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["research", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "55d876613fede9d4611d53ea20e23d38ab40bbcd"} {"id": "train_047506", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["evaluation", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "508a68a32c018907ef856929c644577a8485b631"} {"id": "train_047507", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["security", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "9f35854b1fb3ef5db7c083bd33b73025a5d854bb"} {"id": "train_047508", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "75f205a378bfdf3c2d2177e4fd56fd74476c0e46"} {"id": "train_047509", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Rust", "moe_experts": ["performance", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "5ef24bb2509397538d056f4fe5b714cfb2a7b43b"} {"id": "train_047510", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "f83fa4fad60c9f33c981891468aad8a87b9a34c3"} {"id": "train_047511", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "7ab3315035a7d411958e94c59179431b8f1e39ee"} {"id": "train_047512", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "1f6711bd586c9e37ff7f1ade04fc69215402c12e"} {"id": "train_047513", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "7ba2aca6112ce031342472d5785e99066b536886"} {"id": "train_047514", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["evaluation", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "4ef93dade78d65981a8ac6d4e269d4ab194bebbd"} {"id": "train_047515", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "4178464ee52a77f5d904e7bd478bde6f20f5f649"} {"id": "train_047516", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "6bc7bda13feba1e65833258bc9941917f0e497be"} {"id": "train_047517", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["governance", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "940d788056fdb513c5e4d501d877bc6f99dff983"} {"id": "train_047518", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["security", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "79274bed166f7221c6c3f986ed0ec02539fc5fb9"} {"id": "train_047519", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "7cec732e5d0eba706d68ef89b24bd27a3aa33517"} {"id": "train_047520", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["security", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "feb889934b884a9df6c00d2aebafd283ac4f3b12"} {"id": "train_047521", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "35518f81071cc96ad4bc81c3fcfdb494d5a7aead"} {"id": "train_047522", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["evaluation", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "3860bfd8641d09ff68f6cc68f2f9b21e7178d300"} {"id": "train_047523", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Bash", "moe_experts": ["governance", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "e9d3eeeb9549715841ff508c31adc4528d2005df"} {"id": "train_047524", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "0f29c651bccad8087515c07e1271d77524f01f54"} {"id": "train_047525", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["research", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "8ec4b3093c1557af91e42e463919b69c52dbbb4d"} {"id": "train_047526", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["governance", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "757a7bfc71a1fd64aefabf8e9f6676fd1c22fc0a"} {"id": "train_047527", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["agent", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "cadae7023a5315166b7e18d744542c58152e544a"} {"id": "train_047528", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 9,\n \"confidence\": 0.9,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["agent", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "9ab679e866c47bca6adc979cfc32a9fc41bd275e"} {"id": "train_047529", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "c15d22d2eb0a4bbe877f0be52db6e4b5c944f6f3"} {"id": "train_047530", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["data", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "1f224516a16eec82988242350036be4555426694"} {"id": "train_047531", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["research", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "4b7ec8537200358fb69577fb16ea23b108649a67"} {"id": "train_047532", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["data", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "1fc0c6c1e2d3f37e762f0d926be085ad928f583a"} {"id": "train_047533", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["data", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "5d87870bf356d9df889bd5100af2c818ea719a73"} {"id": "train_047534", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "9b588c7a7155904fa06718139210bcafd7fd64dd"} {"id": "train_047535", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "e1095cba53f9dca58b551d1380ef779a49cd9fe8"} {"id": "train_047536", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n cost = 0\n for _ in range(budget):\n p = plan(issue)\n patch = edit(issue, p)\n ok, _ = test(patch)\n cost += 1\n if ok:\n return patch, cost\n raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["security", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "3928e43ed5968d2ee9b47bd55d972501da213549"} {"id": "train_047537", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["coding", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "4e5e63e36851d02993f1c85307d40583ae6bc43f"} {"id": "train_047538", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "6b93f4d39f4bfd51ee8a6022920f8c47fb50c7aa"} {"id": "train_047539", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 9,\n \"confidence\": 0.9,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["performance", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "25879f72ccd08313ded5e6d98945684b1084fb44"} {"id": "train_047540", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["agent", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "f76ebc82f40d9a4c7fadeb92e4d2ee0f5605f034"} {"id": "train_047541", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["data", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "0b248d66f7d8b64fec02cf2913e8a8a269d4a3d6"} {"id": "train_047542", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["security", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "a30c93782e34a53a1a6725c441548f8bac5d04cc"} {"id": "train_047543", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "6e08e692bda81c92588e587e67e4b6c3e3045f3f"} {"id": "train_047544", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "2baa590ccd02d557a27d4708308bb21bca0dd277"} {"id": "train_047545", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Python", "moe_experts": ["security", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "4baade02d402298f31ce0f6ee7881344a2338639"} {"id": "train_047546", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "1542c31844528484a6bebb1e11230b2f0aafc777"} {"id": "train_047547", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["security", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "d99d9632f7a0611348b28c06f96f97e05e6c89ca"} {"id": "train_047548", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "046e73f9591a3d975af71d9c856f8dead69b7c89"} {"id": "train_047549", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "ab852b2db3058cf21b48d51be264e74c06e2bc88"} {"id": "train_047550", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "b17ebbeb1e8f571cdc634d3e951a2329b5e3f0cb"} {"id": "train_047551", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Python", "moe_experts": ["research", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "c81ecfe17c0907b2e620c97b16f01327291ad7f3"} {"id": "train_047552", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "11522dda8aae6e31bdb3ccfbde0797a48e1090a0"} {"id": "train_047553", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "28d566a3a925b295e0b27f35c4e3adb8f43b18e1"} {"id": "train_047554", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["governance", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "f230463b196dfc7a29234a13cf4d0c2fd742b75a"} {"id": "train_047555", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["coding", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "55ead3871ea2cf377f9cbfdeced50cd0b6e8bbb8"} {"id": "train_047556", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["data", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "4045a6741aa3454df5e515982f15fe9185b9819d"} {"id": "train_047557", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["security", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "bdc384b230090c50841f7b05aae667bb85e138f1"} {"id": "train_047558", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n cost = 0\n for _ in range(budget):\n p = plan(issue)\n patch = edit(issue, p)\n ok, _ = test(patch)\n cost += 1\n if ok:\n return patch, cost\n raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["data", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "0a04f134bc99df20ae1517a4efe46c9f0e991950"} {"id": "train_047559", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["coding", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "79dc26879904aff85a5fafba142965ad41a2baa7"} {"id": "train_047560", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["data", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "1794833462fc090696c87076a3f8267fa447178f"} {"id": "train_047561", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "eb6535e7554551a537cd0ade285351e8a12e852f"} {"id": "train_047562", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "66f51f2946c149779630259b17e051877b7cd4e5"} {"id": "train_047563", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["data", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "94ec4202f4f02ae5b45a15b65cdbc1b46cde2de4"} {"id": "train_047564", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["research", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "4112a0fbbd574836c8b549ceef67af59da9bb79d"} {"id": "train_047565", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["coding", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "e0100b1a6c246288a06a170f332ba469d1e985ca"} {"id": "train_047566", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "90aeb892f9be8f0739478cff8ae790c163038c2b"} {"id": "train_047567", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "eca07127a2934e179f6b9ba212293eeef5ac75aa"} {"id": "train_047568", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Go", "moe_experts": ["agent", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "f60e42b0ed946bd0f0ecd045ee3e37546a530916"} {"id": "train_047569", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "C#", "moe_experts": ["coding", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "3bc361eaea596d74166bd450ff90e317f2b72dd1"} {"id": "train_047570", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["coding", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "e46449e4a4df9b7c405455f95c6e1afa4b1461de"} {"id": "train_047571", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["evaluation", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "729b801c3dcfaa7022040939a1a2db1fe1388c3f"} {"id": "train_047572", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["evaluation", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "8a45c046a31a15052d95e187662198b7fef6ca04"} {"id": "train_047573", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n cost = 0\n for _ in range(budget):\n p = plan(issue)\n patch = edit(issue, p)\n ok, _ = test(patch)\n cost += 1\n if ok:\n return patch, cost\n raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["governance", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "0909b54e25e9724de69788e37130d757d01bdac4"} {"id": "train_047574", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["coding", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "3b2de128a1eb5f41290fe8e629d84cd6860cd8cf"} {"id": "train_047575", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Java", "moe_experts": ["performance", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "b28744a4c20fe7a8aae3a1752456827ec0acc1b8"} {"id": "train_047576", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "e7a5e3fd8509ae6829fbcdb96152e58bbf0b8ea1"} {"id": "train_047577", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["governance", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "61e4df400a84e02fd956d22c50fa4b83ab52710c"} {"id": "train_047578", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "244ebfd72762d0f12fc35f61936f2de69e9f8279"} {"id": "train_047579", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "402dcac0b570ad34a70aedee9992d17af6307758"} {"id": "train_047580", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "cfbca6aeaf20f5cbfa819f8e701718b53b523830"} {"id": "train_047581", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "367d7ba120e55ad8363621d713303dc854465986"} {"id": "train_047582", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["data", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "970b22c3c1eb2888b0412236e0fa4e660f979d6e"} {"id": "train_047583", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["agent", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "d02d4c88b14397b3e0a6e4294a0dbd7bab07dbb8"} {"id": "train_047584", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "13ea882351b0a2d5c157efb0801a35e3e86c9631"} {"id": "train_047585", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "741196d184af649daf132e5fac753784a84b30b5"} {"id": "train_047586", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["performance", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "35d873bcf1486362383ff488265dfe865fc299be"} {"id": "train_047587", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["security", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "00739911e10817bb54722633c95bdc0cf69396ac"} {"id": "train_047588", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["governance", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "10b1f301dde5492bef251a0518d8bac85645f9d4"} {"id": "train_047589", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "5591b3e4b28a3a25e442b89784fab77af9786968"} {"id": "train_047590", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "db05e2ea1ada67ba4363502b6e62485e4a45038f"} {"id": "train_047591", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["research", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "8404805c3d317276d6344fea20427cc34bebebd6"} {"id": "train_047592", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "694db0a9bb0a5ee6da216836b1863a72d9196508"} {"id": "train_047593", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "3a771d21227f8aff5b59cc22aafc050818707e70"} {"id": "train_047594", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["coding", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "85fc53b09d7af8c09567edc87b40621a9f087722"} {"id": "train_047595", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "6f6a418467f8013c11939c8cb1dff37d3b2d5cf2"} {"id": "train_047596", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "ece056665de7e414a0ba0367c38ec6b15009c652"} {"id": "train_047597", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["research", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "816ecf468d299bf6fb2876d055ed465b66809f1a"} {"id": "train_047598", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["agent", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "58f47fb6de6aec04040fa36cdbeaac655d07f465"} {"id": "train_047599", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["security", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "319741d31a105ca388d484b2d0927f3eee7ab66a"} {"id": "train_047600", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["performance", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "641b04142ea40fab0976606175e70128e083a280"} {"id": "train_047601", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["evaluation", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "e42389fde0d306edc2a9ac494df1094cc3d1933d"} {"id": "train_047602", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["performance", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "4b5e1568eecdc98f5abc5c6c0ed202675eaf2565"} {"id": "train_047603", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "98d813860002845d670e83d810902f45953009d0"} {"id": "train_047604", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "a71e86bd47de2cb59299052465abeeade7907052"} {"id": "train_047605", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "8cd2faee073faa53efe1129d20e4301a92f96a00"} {"id": "train_047606", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["performance", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "cb13173795d6828539cc3ad3bf86e0582b0dcce9"} {"id": "train_047607", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["security", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "1179ffb82e2fca3cce6c47eebd448f33b2a6b6bc"} {"id": "train_047608", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "c0ebbb30825144ab846789d6446ad102e20688d6"} {"id": "train_047609", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["security", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "afa1b36b47519fbf910623732daa454876f6d2b8"} {"id": "train_047610", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "C#", "moe_experts": ["governance", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "bfa670b9c2978994cb79a13a6d005cd0047c3355"} {"id": "train_047611", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["governance", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "ec9412579942a798bc8985c5ff9da0c5488b5c16"} {"id": "train_047612", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["coding", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "4f195faf2b2d6384b4ef4f9e166e18f367a7e0af"} {"id": "train_047613", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["data", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "36c70049968ac00fd297e4eff25f0ef4e832be61"} {"id": "train_047614", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "df91b2ea749e3c9f6bc48739da5fd10c07a3599a"} {"id": "train_047615", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "501f66c4fa45b17afb3c438deafafc9d0e57dace"} {"id": "train_047616", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["data", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "0bc58ec4d7d90f206a0b10697ea87af263262c67"} {"id": "train_047617", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "12d55912e3b3ae6b857d279e32fad6962b4aefa3"} {"id": "train_047618", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["performance", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "63f5fc8e110c2651d97ad904b58c1d07aa32e574"} {"id": "train_047619", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["performance", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "ffad3061c20e95198f0cf38f84b7a83f1523711c"} {"id": "train_047620", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["agent", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "cf91af3790dd2ae809e608ce6cf58d04531e4a86"} {"id": "train_047621", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["agent", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "0d3a52bb9aa15610f3a7108d6978b973906ea8f4"} {"id": "train_047622", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["governance", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "2eae75e7b4edba4ea9c09e389e71f0bc48df1eeb"} {"id": "train_047623", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "136bb8fcc51a9a6e3f38ef0f7950fc31e518de55"} {"id": "train_047624", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "7ae0c9ac15f3d2bc9ed972a6d70bb338f9df1d49"} {"id": "train_047625", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["evaluation", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "cf06af8502a953faf36efc9c07f492b5c4aaa19b"} {"id": "train_047626", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["governance", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "672be8229d4342f8a45111dff4092b34ce5271c7"} {"id": "train_047627", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["data", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "e92c5b0fd22f29a1318379ad020e4698c9d1f1ef"} {"id": "train_047628", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["security", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "f59162e0ec84f91e269b3fde76b14c051af21b4b"} {"id": "train_047629", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "e5b3cf1cf52601a6b57383bec9f37c31a51b98d4"} {"id": "train_047630", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "54a39e664aa04b47595dde6b02947bb5d9bb3e39"} {"id": "train_047631", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "SQL", "moe_experts": ["performance", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "aca413a18faa722b2b8beff812b1758572b4c2b7"} {"id": "train_047632", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "ff5c4397e404f66861ed8c042b9891a01e0f7ef3"} {"id": "train_047633", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["performance", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "483319a9808d023927b09d94bb9ede9f78961492"} {"id": "train_047634", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["evaluation", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "55e586ad38ea4ad054d738846fc4fd2b1aac4d7e"} {"id": "train_047635", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["security", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "f34e6d9c53f59f2340b9f7ec3e136f708e15be5e"} {"id": "train_047636", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["data", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "f30cdbf1278246c6a063a43937f8328ef28abf0c"} {"id": "train_047637", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "6f60bb6074a9d4e919ff53cd216b3be6a2f9726a"} {"id": "train_047638", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "b1e2d7c2361adb53c5803a68ca3031a47fbace0d"} {"id": "train_047639", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["research", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "79ac06d9c855eb91613c2c4824f9d30fb6c09626"} {"id": "train_047640", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "75dc639526f675ab1b257fed87fb5f4e4ed867f2"} {"id": "train_047641", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["research", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "ee4b99e496711962a25314f4bf4544d930c7c2b1"} {"id": "train_047642", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["governance", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "2c62b8b71bc14231aaf48703579c4cf74c6478b9"} {"id": "train_047643", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "a0a4e92220dd9fac9837c006ccbfe8ea9dd4ec5c"} {"id": "train_047644", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["data", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "e1ab0cd71d20280fbfb7659ce9d52edf93fd6390"} {"id": "train_047645", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["performance", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "ff4bab78e80c9a9d9bd830443d2b123386cce057"} {"id": "train_047646", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "fbebf3c0db9870560b681eb362112ccb89b97024"} {"id": "train_047647", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["coding", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "c88bf5c6f234c6eaf24eb988e13039c14529ea66"} {"id": "train_047648", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "c8a38d18110930dcad05e08901496258b4e79b30"} {"id": "train_047649", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "a665980123735676992b568e7bb08b05976c5732"} {"id": "train_047650", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "f597b21f4124c3354007c102eb4b45d37c65057d"} {"id": "train_047651", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "2ab17f50df5ee830d501d849dd9d3f6f43c3a24d"} {"id": "train_047652", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n cost = 0\n for _ in range(budget):\n p = plan(issue)\n patch = edit(issue, p)\n ok, _ = test(patch)\n cost += 1\n if ok:\n return patch, cost\n raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["agent", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "4607d6d5357135e047fc12666effd1514887df97"} {"id": "train_047653", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "ad36866f94ee445abd719037fa5ff7f4832dc5cf"} {"id": "train_047654", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "0493ad517c86b096548eed241df4c3524bf85442"} {"id": "train_047655", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["governance", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "a5d7e47ce92a611f33e0e84f6a7478a62b672813"} {"id": "train_047656", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["research", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "3549bf7deddd27a2df06e0f062209169ebdecbb7"} {"id": "train_047657", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "2b4fd7c5b917e28ea9e4910d0e44eddf37b4e692"} {"id": "train_047658", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "fddbc3c4e3874b8c074e2270766eccaf326b2abc"} {"id": "train_047659", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["data", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "4f4befc7e55932b5bd0b6469f6a05d960f297bfd"} {"id": "train_047660", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Rust", "moe_experts": ["agent", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "e3cea6e1a79ddfffcf9a05994cb16e6997d4f265"} {"id": "train_047661", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "0690e629dd3fe259862a63dbb3cb156f6d63072e"} {"id": "train_047662", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "3cfd24122f9d9da1dd62681ae904c2a5161c69ff"} {"id": "train_047663", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Bash", "moe_experts": ["security", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "cc6e3f41dd8b5440e60d965a7111de362666beb4"} {"id": "train_047664", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "e6ce236eac5383cd280343aa2ea0bfb5eab1a45f"} {"id": "train_047665", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["security", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "9da05596451bb1d23fcbec560a54aee3cc102669"} {"id": "train_047666", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "50efcf88b32b862b3e4764d804b1da467473f3d4"} {"id": "train_047667", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["evaluation", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "6d3872d7afaf8fbcf82b98d8ba7a77ee63024e4c"} {"id": "train_047668", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "6719b528e5df6966829edc5624295ea0fc9435a1"} {"id": "train_047669", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["data", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "2517075c8d30d52b71d86b53e04e79e1a97d2e1a"} {"id": "train_047670", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "e539a837717c3a39c07fa6dd09f1f74ca706d850"} {"id": "train_047671", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Python", "moe_experts": ["security", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "baa3546f83f13f013677a56192753d256a6b9244"} {"id": "train_047672", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["performance", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "6a44646789349f01ebce65abc15eea70cdde27dd"} {"id": "train_047673", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "39c757fb9e79aed888e50311497400b05d3145ac"} {"id": "train_047674", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["security", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "df856c34b8acd446b0188b4847f43fa0830505d1"} {"id": "train_047675", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["governance", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "7dbd02ce26341375e07d7e8f9b4f059342296f8a"} {"id": "train_047676", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "861d402cd44118c07b3c6cd0145a725e7a2735eb"} {"id": "train_047677", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["research", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "8ec4b3093c1557af91e42e463919b69c52dbbb4d"} {"id": "train_047678", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["governance", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "b1d6112cdbdd005a0dfad833f42d75f8534610af"} {"id": "train_047679", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "5c6c5393934e203672e50b7046a528a3d8b6239d"} {"id": "train_047680", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "f6c8c0c779492436c1af5b2df7900be59c180e23"} {"id": "train_047681", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["governance", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "0cc615f989e17d797951782cab4d936e926fe873"} {"id": "train_047682", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n cost = 0\n for _ in range(budget):\n p = plan(issue)\n patch = edit(issue, p)\n ok, _ = test(patch)\n cost += 1\n if ok:\n return patch, cost\n raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["security", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "2c27c2dbe176249eca18a1ba70cd0def909a5f0d"} {"id": "train_047683", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["research", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "f032e42b800f15b1d45d0331bdbf6fd8054172a1"} {"id": "train_047684", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "fc50f527cd983137614dc677692d954e1550d54d"} {"id": "train_047685", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "6cce94668f3f644857ea18bac2cc9b1796159685"} {"id": "train_047686", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "cf1a2404771bac3134ffe20926cdf6f59e3fde89"} {"id": "train_047687", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "SQL", "moe_experts": ["research", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "fc73a23fa58ba71e6749840dbb92e13c7aad4907"} {"id": "train_047688", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "da8b38d463fe6fd33a480b22a113143adde8d7ba"} {"id": "train_047689", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["performance", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "5f52fb38d3aca202fe634087304b751c3eb473eb"} {"id": "train_047690", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "10d9773fbf00f99974991f060bccecece925bb38"} {"id": "train_047691", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "73bdc2620982e29d0d6e328a05ac81472ab79cb9"} {"id": "train_047692", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "8b5988304797ffc966e7e66e494f6fefbc2b7d0a"} {"id": "train_047693", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "4ff809520b8b0575e82966420ac0c0fe5e6018ff"} {"id": "train_047694", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["coding", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "57fee23f63d9f4cf7d6c34a1db8b748b7465b54f"} {"id": "train_047695", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["agent", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "69b12a3766eac5ab54b019bd585b13c26e3bc9d7"} {"id": "train_047696", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["governance", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "b9ffb2636bb4ca2bd880262ddc39c4be1357bea3"} {"id": "train_047697", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["research", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "693567cbd616862f8134258e13d0bd9ac40483bd"} {"id": "train_047698", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["coding", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "80487feef9b272b7f8aac9330a068c2c2435addb"} {"id": "train_047699", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "9076202dbb276537389cc5a038c447862ed4b819"} {"id": "train_047700", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["data", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "93e87cd216e5005cc3d4a0f6e82bd821a6217338"} {"id": "train_047701", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "b17217cedb67394c5bbdda6835e874e381df70ad"} {"id": "train_047702", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "05d0d1d9dfa8d9d427de25836719196572c1d1c0"} {"id": "train_047703", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["security", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "c2ff70bc1c141d48cd4ccbd5bf74a8028e2aa37c"} {"id": "train_047704", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["performance", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "df8ea1372eb6fec5af4dc90e95dc4b48f9fa0887"} {"id": "train_047705", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["security", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "458f14dcd883e04df43989012b0c43d2322795f1"} {"id": "train_047706", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "2081053be27b5b2c6bcff63f0a55994d0e5243d9"} {"id": "train_047707", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["security", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "dae467afb2fa6364c0f02dbc2cb929dec176d18c"} {"id": "train_047708", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["security", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "7700aea648a78780f5b1c78bcbc75931401e0349"} {"id": "train_047709", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["research", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "d8847f03c0800f45b6b7d5eaa315540a131905a8"} {"id": "train_047710", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "a992bb19ea527b99b3f51fac304583cb5def0ad8"} {"id": "train_047711", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Rust", "moe_experts": ["data", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "e3cea6e1a79ddfffcf9a05994cb16e6997d4f265"} {"id": "train_047712", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "c5a8c190d5bb177e147c27c8d05bce3ece7743ac"} {"id": "train_047713", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["evaluation", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "37a46ba1c8c584a6d77e3b0dcba5b106a801e560"} {"id": "train_047714", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["data", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "13447e93ae0b693d2d81a2959f400b6b85a22f5c"} {"id": "train_047715", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["data", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "458f14dcd883e04df43989012b0c43d2322795f1"} {"id": "train_047716", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "132927888e54a1e125431ab7faf3ce0047cf7570"} {"id": "train_047717", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "6cdd0f82ffa11fa9cbc68ad13bea59505edb5f1b"} {"id": "train_047718", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["evaluation", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "2a7dc411f37f3cabe98e65c1dded8c221c3d4f22"} {"id": "train_047719", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "9f4ba7aec51ccf7ea5a3fcb7576edaccdb1d8af8"} {"id": "train_047720", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "289dd6997f901a6985c2cb8686a54a76ca2e0f33"} {"id": "train_047721", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["data", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "a300be8952907cf69609cd8c38c4ab0675ba20a5"} {"id": "train_047722", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["security", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "3bff15f03f20202bdb26efdc979cbbe51bb1b3e7"} {"id": "train_047723", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["performance", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "58248f30d04c57ce1f9d1a5397d17ba7fb86f5d2"} {"id": "train_047724", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "ba9afdbe55388a173fe7b56aed8867b66c10ed0b"} {"id": "train_047725", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["governance", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "4983321e57466d045d21f1d285519b5d12ae47bd"} {"id": "train_047726", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["data", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "8d9d2d24abdf50f827584a0f2134d5ee6c63ce9f"} {"id": "train_047727", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "38f1a0abe1ac4fd08c77a2127de90867e225a4c0"} {"id": "train_047728", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["research", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "6e2e48f5a1c640b6881f2dc0797ae5f1d7ef0a11"} {"id": "train_047729", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "1a2cea4a5ed951572bb49cc8f04c0c1be186f1ff"} {"id": "train_047730", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["security", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "c84e420c2afc3e010834f9c3885d28f1d46a6482"} {"id": "train_047731", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["agent", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "e016028a128c1646ec2e5102e26249e198f7c11e"} {"id": "train_047732", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["performance", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "c57454a4da244067f0c863750e23859003e61124"} {"id": "train_047733", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "8e5820bd57d0c79c88b21e454961a9afb1d8de24"} {"id": "train_047734", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["agent", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "2cdea0b3b643340e729e8bbb722b4d1332a20572"} {"id": "train_047735", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "SQL", "moe_experts": ["agent", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "5207ad9d7df02ac1e0bf5cb7daeecaae031ca41d"} {"id": "train_047736", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Python", "moe_experts": ["security", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "baa3546f83f13f013677a56192753d256a6b9244"} {"id": "train_047737", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "354f87088d12f7dab8ac7389c5cb78aeb592ed05"} {"id": "train_047738", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "9d916793f091bc28e162b332c0ce112ff15dcf95"} {"id": "train_047739", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["evaluation", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "a882b703f279444012322c978df12bf692dc6a92"} {"id": "train_047740", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["research", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "6919563a15a46e84ef60df9097c7cce096613ab2"} {"id": "train_047741", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "7600a837f6096844430c6ff265bafb2c87dd71e8"} {"id": "train_047742", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "f7e5b275e6f89e0dacb069e93b4e6ae13815ecbb"} {"id": "train_047743", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "92b3068096725b829c5e027c59b589826ced7b50"} {"id": "train_047744", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "35d957bc34704b1d4ce1ea6dbcee644964dca3c9"} {"id": "train_047745", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "8d5fb3468c4c1a507edfdb62b289402b0712c9e7"} {"id": "train_047746", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Python", "moe_experts": ["performance", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "e0a25735124a3c77bb37d05a9934602e7998b460"} {"id": "train_047747", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "47d5e3924f1613372e402987887d81de1499ef74"} {"id": "train_047748", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "7e8d4b070707aad57cd4348c18aecf85de27e9a2"} {"id": "train_047749", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["security", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "35d11af3d4d896444a088822c040eca4d3bf5b67"} {"id": "train_047750", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "e9c23641b8a773a4602c83c6597c8d03e47b1cc0"} {"id": "train_047751", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["governance", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "22d27d1120b9ca3b5ec07a5bd21ef6bf5bc639ba"} {"id": "train_047752", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["agent", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "2d768d959a4338bf2071f61935fda646a83b1dc9"} {"id": "train_047753", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "37457e2d679284e2402b287ca489a2eec26f562c"} {"id": "train_047754", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["data", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "1b66af898feeb99326b6f21c8580b5718855596f"} {"id": "train_047755", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["governance", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "badbaed4cbbd204ddc134abc2a6d3f285d1ae5bf"} {"id": "train_047756", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["research", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "3cbb2ff0c24ec5426d8922ec13c47b211b4f77a7"} {"id": "train_047757", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["data", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "23fda4cb48c34e2a4b50524501627add28a8079f"} {"id": "train_047758", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Bash", "moe_experts": ["research", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "fe130f695827b856942d81033e319f0e0c54e0a4"} {"id": "train_047759", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["agent", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "4d37d115ca988258f6566efc44ee0dfadf2a4612"} {"id": "train_047760", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["coding", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "5233288905d1ee477c217b61273c3f85f00ba1d6"} {"id": "train_047761", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["governance", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "a2b04111458c333ad8c1bd8c3921819a93f36495"} {"id": "train_047762", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "0ec788ac0040599ec7aa6192d93a9b9de24127fe"} {"id": "train_047763", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "3c2f9f7a8713bdf6a6c303a0c664e4714f70e0c8"} {"id": "train_047764", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["agent", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "9efe89c01efc062c4f9da16bde9063ebd868ecbf"} {"id": "train_047765", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["performance", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "11c1d09e71b4877f31d4c13e8a4521886d95c34b"} {"id": "train_047766", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "a6cc73a86f983ff9ec14cad436db506b1c790bc0"} {"id": "train_047767", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "178548142c851fc6febb65768198d94b6d69fc87"} {"id": "train_047768", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "efa4e123044ca400bb4fd3179f4e539bbf142dc2"} {"id": "train_047769", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "f8d5f3c0581824a9f96c2a4a85501d4682cf4e40"} {"id": "train_047770", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Go", "moe_experts": ["governance", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "4b0509be2b707ab199651c03aaba063e0674cb63"} {"id": "train_047771", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["data", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "2f0425f1bd64ab2dbdb3d0364eafec8a59a1f223"} {"id": "train_047772", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["governance", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "1030c0f689be5094937b5df0a31e247c45240e05"} {"id": "train_047773", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["data", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "41a12146278e1e605b416abb60957dcbcb7c85e5"} {"id": "train_047774", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["research", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "8404805c3d317276d6344fea20427cc34bebebd6"} {"id": "train_047775", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "843dbed2f0274440416ef79452031bb94bcb0850"} {"id": "train_047776", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "84efb9d2ebaff04b0adbca600392f726b488d161"} {"id": "train_047777", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["security", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "5ab5d8ffe83b7f7bd954ab516e199e097b964584"} {"id": "train_047778", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["coding", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "6edc00799b28e0ca20c570020a4f93f855658bf4"} {"id": "train_047779", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["research", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "a48faff4296c3872e45e930279d4c51b03297ac7"} {"id": "train_047780", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "b3a28948ac1c950c38eb82197275b141dcfc6be8"} {"id": "train_047781", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["security", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "ff6ead12fdc96534d2c6c91c543d73b6e6fc5659"} {"id": "train_047782", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Python", "moe_experts": ["evaluation", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "2b531a4b672ab0479f748722464f4ed16a9aae8d"} {"id": "train_047783", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "23b8cd57d7e38f1140a2395ce0da2e58ee9261de"} {"id": "train_047784", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["evaluation", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "b93c2eaa45a7597e49eed196107305d5cf9f7500"} {"id": "train_047785", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["research", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "6841b535e4293e04edb9262b073a1ebc9392dcbe"} {"id": "train_047786", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["coding", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "f6a2c092c7c86c79daa8aec79089523dbd3bcc97"} {"id": "train_047787", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["research", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "7785ea2e8277b140ddc68bb39de684c2bfed3a99"} {"id": "train_047788", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "53501ca9af9395c99028258f5a3fa6553e1acc37"} {"id": "train_047789", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["agent", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "8bde51d5438e78b392290a76096460c365b80cda"} {"id": "train_047790", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["coding", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "ba64e69b8299d009a7adc757da07fbaa626d2422"} {"id": "train_047791", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["data", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "7c95154192feee57495e042b4fadda4879b2a066"} {"id": "train_047792", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "36d325a04986084b2b905af83fa00017e8afb9d7"} {"id": "train_047793", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["research", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "4e6689af1cbad5e1840ed562e05161cf0db51d39"} {"id": "train_047794", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "f28f611e1a4bdfc9951d74914138d1c6ebc90f34"} {"id": "train_047795", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["evaluation", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "3c3b89e9be57a31d42372f0a9996afe792bd765d"} {"id": "train_047796", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Java", "moe_experts": ["governance", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "b1ac1fc910a3b4f3a98722b55214550c16d42710"} {"id": "train_047797", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "f5b86838b5f8eb264b88fb0728fdbdd925878ac8"} {"id": "train_047798", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["agent", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "67a5e99bb941f3f6ea1e7604a35a12094eea8c0a"} {"id": "train_047799", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "5cfe8993db61dae9b824d2ba30e893c73a89df57"} {"id": "train_047800", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "dc46fc710b9b6726d1dfef779baf0e37c2b871db"} {"id": "train_047801", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["coding", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "3e78e9e1a978c438aa22f7b89206c9090890932c"} {"id": "train_047802", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "7aeaf9e9fafb5b1d883926d80b060e410447a6f7"} {"id": "train_047803", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["coding", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "4e861a3518a29cf86d8fa457fbcb4fd0404e0c74"} {"id": "train_047804", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "8f90cc6e45fd065f3872f7dad0c45221589323b3"} {"id": "train_047805", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "ceb518842f8a3a44e475c535ba0b97530d6894ad"} {"id": "train_047806", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "9b81bf64b6ae1f0f7415614f9d2b69a15448a6d7"} {"id": "train_047807", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 9,\n \"confidence\": 0.9,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["performance", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "4c017aeee98c95d2dd8965f9330427f0800417d7"} {"id": "train_047808", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "984daabc6a61b868f56df7c80ffa088437e086b8"} {"id": "train_047809", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "8f4904588fa27d8ab54b9ee15ec00549c5f59da0"} {"id": "train_047810", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["performance", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "01b16b75482c36e37468d97c217e91c6b9ea0634"} {"id": "train_047811", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["security", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "22ebe33bfc1989a7bfd452bc24395fc2c4b1d34c"} {"id": "train_047812", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["research", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "1591c8d42b4bc99b80252843d79418e6596a5972"} {"id": "train_047813", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "f2983b3902b1c13871da277f75cba40f14943d52"} {"id": "train_047814", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["governance", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "f3b259a3691623114b7ef65c63ec6cb2f8907b78"} {"id": "train_047815", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["evaluation", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "b07e58856d37da2b11ccfacf071941896a4b8fa1"} {"id": "train_047816", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["agent", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "b473ee2b911add6c4d262f69780c5b1231319809"} {"id": "train_047817", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["coding", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "06589a2c9ea8475873e5db4c2c6258b310b68087"} {"id": "train_047818", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "ba24666a7c93bd9b6d3434ab68a30be45688d1ab"} {"id": "train_047819", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "701ad2c629f555aa0f0b5927158d856c96ae3109"} {"id": "train_047820", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["coding", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "7f9ca58df29f3dfabd8a86dab4aecd489e9b6b55"} {"id": "train_047821", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["data", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "b35a2dcc23055919149a00002f8e7995f8fdce92"} {"id": "train_047822", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "C#", "moe_experts": ["research", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "bf563d64b3183f69ebda641bd4e39524fdac97af"} {"id": "train_047823", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["performance", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "5ddf61c2795cb0b8d548a70079f25304758b64ba"} {"id": "train_047824", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["security", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "e9c15abc9eefe55c266945f616f50d76961ec359"} {"id": "train_047825", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["agent", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "4de9016f549e5e884a6496d2d83761f890a58efe"} {"id": "train_047826", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["research", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "728bcafb345381d32094e4fcc901490250804379"} {"id": "train_047827", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["performance", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "4f1705d3360c2d2209b300c11de919e3a4b5051e"} {"id": "train_047828", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["governance", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "3332cb2f6d470424f2dda572b42edb68b7f2f983"} {"id": "train_047829", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["evaluation", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "9bf4d79f94061e4e784061bd386d298ec2e80802"} {"id": "train_047830", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["data", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "2e3354489ee22720deb7638a417931196aa32b59"} {"id": "train_047831", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "22ce7ef0ead00525f591d26081b2bab1fe470da7"} {"id": "train_047832", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "8a62e2b91b9ec02a1e5258c22a935b952d899458"} {"id": "train_047833", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["coding", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "5233288905d1ee477c217b61273c3f85f00ba1d6"} {"id": "train_047834", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "SQL", "moe_experts": ["coding", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "f9c9002ea90aba6a480c4b6f9a27b9b4c35941a0"} {"id": "train_047835", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["security", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "3224125a421679abb5664020ef6f76d1e150ae5e"} {"id": "train_047836", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "55a7a551073937f2ed8aa304987ea29dcff5354c"} {"id": "train_047837", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n cost = 0\n for _ in range(budget):\n p = plan(issue)\n patch = edit(issue, p)\n ok, _ = test(patch)\n cost += 1\n if ok:\n return patch, cost\n raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["performance", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "424e37c8a4c0ef131d37e6f5c3b548cef8b087fd"} {"id": "train_047838", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["data", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "c34e8e85e2f1a418d683c61854d65e858efaa25b"} {"id": "train_047839", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["governance", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "0d212f38df7fe9d9b3498b98da710201b05c34d2"} {"id": "train_047840", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["data", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "0c7fc467367c957b711343eb47954707b3c258c6"} {"id": "train_047841", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["agent", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "4d1fce3ad3829aabfa003878207515a2a2f54b65"} {"id": "train_047842", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "91f874d9323fd932e4e8f9dfed0398cf6a993c8b"} {"id": "train_047843", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "0b98414af0cd603582327b1d91917b4ddfe92bdd"} {"id": "train_047844", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "5e24b084a772abdb4142e93b877786ad953f79a1"} {"id": "train_047845", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "f348728ba405707257d1232e8927ea7594c64c6e"} {"id": "train_047846", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "156147a93484a2ffd9cce98a6bfce1ee816762fc"} {"id": "train_047847", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "2cce4b83360a7511c48594fd82fdf9af00943dd3"} {"id": "train_047848", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "0afab87473e921d1eb85dec68555c0a3b807f5bd"} {"id": "train_047849", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["research", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "6404f737f7fa8f7dfb969bb4822ba440f0bd72eb"} {"id": "train_047850", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "fe27b85a0426f61c74f84a9225f8581cb0a0dff0"} {"id": "train_047851", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "2338ef8415a66ad1fddb23b2d4c570b2dd5a5a52"} {"id": "train_047852", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "e8b997c5f7e711e60b4bb1e7c0a2c4d7f9ba40f8"} {"id": "train_047853", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "498a9e7f483ba10db79fb3b6582e9b9be3c5864c"} {"id": "train_047854", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "a81c128d2590018559a61fb08c30abe1655b57b6"} {"id": "train_047855", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "90dee186f8326d8b0cfc673401d84acb59fea014"} {"id": "train_047856", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["security", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "4d6c5c31881061cb60ba1c200f7218c0c9c28fde"} {"id": "train_047857", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["agent", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "7f4607295d8b7b6b8a7df96b919178a651e128b6"} {"id": "train_047858", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["data", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "159b3d66bf0c0ec518f79743ccb58c8af4c10a91"} {"id": "train_047859", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 9,\n \"confidence\": 0.9,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["performance", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "a0b42cf969a6bd2c96fa0f170f992f406101a661"} {"id": "train_047860", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Go", "moe_experts": ["evaluation", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "7b1e9b3775af3cc7b96dd9cb5a15ffba35ae824f"} {"id": "train_047861", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "73e6657c630cf2ef65e46d6950ea096ae211bb79"} {"id": "train_047862", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "021b6817e5323056a4be43e92d6a9c14c35529f9"} {"id": "train_047863", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "2f3f97958568ee58b245dbe80e7a5a60d8734347"} {"id": "train_047864", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["coding", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "515d3301fb27dd4a0571dd1511ca70a270c08ce9"} {"id": "train_047865", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["data", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "c99d7cbabc0e954cba2a95fe501c5415c155f30a"} {"id": "train_047866", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "8ed6799c37d457ff8e2c57180be3e1b1d61920d2"} {"id": "train_047867", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["security", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "752166c09b882500e15712b5fd2e4cb88dc8f0a7"} {"id": "train_047868", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["agent", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "7081f9b5811b79cd1ba6853ad375b54d3f43f0c4"} {"id": "train_047869", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "33165263f71935f77359e2fd014c12616db926c0"} {"id": "train_047870", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Rust", "moe_experts": ["coding", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "7c90998d0030a3e0e4ba65f6c0706a342a8b45e6"} {"id": "train_047871", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["governance", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "e259007daaeff209655608019e1727b7ceca2de8"} {"id": "train_047872", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["coding", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "cfc6250dfd71606799de10eef1516346007d2ae0"} {"id": "train_047873", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "026eaab5afdde359f87ed0fc45776c3186192290"} {"id": "train_047874", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["data", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "65545a6514b51d400727a600c79a40bcc0e7454f"} {"id": "train_047875", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["coding", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "d7c3a2582823794d3e21938ac2861268c8a9855d"} {"id": "train_047876", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "3b98c96aab3d4d2ac54f911e753cfd2e39e4a9f1"} {"id": "train_047877", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["data", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "967f1a7ae4581cbfb0629beeb5afca7bbbc4a496"} {"id": "train_047878", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["governance", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "ecadab1cf90b696fb283bcd65c8532e6f509dbbc"} {"id": "train_047879", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "C#", "moe_experts": ["evaluation", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "ae32c63620d5976494abbf6519427f3f8ce8ef48"} {"id": "train_047880", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["security", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "8dbc74bf8e34b3aebae0a93a5c3dddbcbc7f9ac0"} {"id": "train_047881", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "eb6535e7554551a537cd0ade285351e8a12e852f"} {"id": "train_047882", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["security", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "9523ec4a9cc6fe6abb415a6384171c93f4bdea1d"} {"id": "train_047883", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "19e6c2d16cafa5414cf76560f1ec7c21841adf4b"} {"id": "train_047884", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["governance", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "d7981bd6915be2a969f636cd75d9ec5459a1d7c1"} {"id": "train_047885", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["coding", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "7e43b7029b231015d6eeec1e1a199ceca6826472"} {"id": "train_047886", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["performance", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "f032e42b800f15b1d45d0331bdbf6fd8054172a1"} {"id": "train_047887", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["agent", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "cb2136ccdce9db0aec891e05aca32004627c3c94"} {"id": "train_047888", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["agent", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "49f53647eb63c1af104021c4bf86ed352ae94213"} {"id": "train_047889", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "64285b074e135203b687b3064567815a6d4f5b3a"} {"id": "train_047890", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "67ce30096e2c9dcbf941e8cfb3e78f00c7817fd7"} {"id": "train_047891", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["agent", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "3761767b0432783019148421bd067502ce77e7d0"} {"id": "train_047892", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["governance", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "3afb25b3747afc3c47ee7e333fbcef21bc8ce7ab"} {"id": "train_047893", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "adb5954604d53335b2b4e0911eef7e9bb89f832a"} {"id": "train_047894", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["evaluation", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "e80510af113eccc1dd930803af836df69e1f3038"} {"id": "train_047895", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["coding", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "7c505764ec7cabc9f41b7664db4c189d10e3ade2"} {"id": "train_047896", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["security", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "be1818ed58696c1e96f599a4537cd81d52fd035b"} {"id": "train_047897", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["governance", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "6ded2e3290e9d072571d2455b4e2f9255a804558"} {"id": "train_047898", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "88a2488ad361e534d2bb58e2aec98b5a93f01660"} {"id": "train_047899", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "cebe87c5f1d9c5d765f6eab3e8a711c5ed12e1fd"} {"id": "train_047900", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["governance", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "3c5d63ce00da209a4457229aef831fa492bd73e4"} {"id": "train_047901", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["data", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "3789282035271e8ed43ee147c345ca87481a4988"} {"id": "train_047902", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "5f924332e195a492fa5f78de69d2451a15e93a58"} {"id": "train_047903", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Java", "moe_experts": ["security", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "acc50760a5b91e095026056edc0807c60ac85575"} {"id": "train_047904", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "9b355aa04352046e3a13e1043bf5f0cab2685788"} {"id": "train_047905", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["research", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "087dbe3a56c8c5dc7e4c448c527ac2d86299a1ed"} {"id": "train_047906", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["evaluation", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "ccd65bec1a55fe45a2e15a3e6cf2683bdffb5b0f"} {"id": "train_047907", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["governance", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "3d4768c8629fcbe607dcc7b801ed8ff49fea0987"} {"id": "train_047908", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "cd92fb80ee3f1abb1fb8fccf6f89f985fb3af129"} {"id": "train_047909", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "1cd28719665934b55db7b9c835cb3196529dacda"} {"id": "train_047910", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "a88e02120226abcbbd7f466b3f588420fbd6e3e4"} {"id": "train_047911", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "a9c36c70a9bfb7112245cca72ac177207f861988"} {"id": "train_047912", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "a7135f87034ec0b24418e09ca8683dfdcb89a370"} {"id": "train_047913", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "e46d13800ae5bcf73849108e93202632eae0ee8d"} {"id": "train_047914", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["performance", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "a64f2796365a21052c264ff225f031202cc5dab1"} {"id": "train_047915", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "35751a1aeee18d6b5a0a969c8cba3daba1025e4d"} {"id": "train_047916", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Python", "moe_experts": ["data", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "9a2536f097203421465dd0acbaf4e971c58e6979"} {"id": "train_047917", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["performance", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "48d022763cba42af4ef3a49a972cba9f5182269a"} {"id": "train_047918", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "0a800d6e197ce89ec4ad352588b9761fbf3fe724"} {"id": "train_047919", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "b709d796aa5a5e039b51a790adbf38178fcc484d"} {"id": "train_047920", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "5c508d54da42e6a6e752301d4e95ab69a9e48d8f"} {"id": "train_047921", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "1adae247b6be3d151185333af57f81fde046548d"} {"id": "train_047922", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "d8b83929b9a1edf7203c9bbec349a86a25183d09"} {"id": "train_047923", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "25c1794a594d90db9a8068cb03c4ace10ecd1bb8"} {"id": "train_047924", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["evaluation", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "34569a5522b163f6fa8a75063f0f64ea0937f365"} {"id": "train_047925", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["performance", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "85af49378f23a15ed73038618c181f7f9902c44f"} {"id": "train_047926", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["research", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "8fdaf1e1cde9d491f355560c8924203526adcb06"} {"id": "train_047927", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "9f1cf54555a5fd47f2be9b1df50d8d9c4207dcdc"} {"id": "train_047928", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "3862ddef5dc8ce80c36ed4d9122dc54c7a02ab1b"} {"id": "train_047929", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "c01e2a88423d8c5860ee456e8e64d5bb98c6d3e9"} {"id": "train_047930", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "48f64acfc1b54cc0dada324b6ed40e2b29fdc0ca"} {"id": "train_047931", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Go", "moe_experts": ["coding", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "31a26834740a7591fbff3cea9e2838a111d7d3c8"} {"id": "train_047932", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "c0b3bb456077b3e56a54d57ff101ff3beb6a9455"} {"id": "train_047933", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Bash", "moe_experts": ["governance", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "8a04e0f5ed56c80aa8f0dceb7b61458aac9655ec"} {"id": "train_047934", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["security", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "3e14f5a83e1feb31f568176f20658ce58a1256a8"} {"id": "train_047935", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "31ef3531359c0e2e11211cab360e8f5d0fe7e304"} {"id": "train_047936", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "919f30984f27cdbe9e0e96c26f881fe29539a24b"} {"id": "train_047937", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Bash", "moe_experts": ["performance", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "eb013b484da060ee300cb0c0e42f62f01c290393"} {"id": "train_047938", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "da39230c1c0fc186346a9dd5738ee6fd20b986f8"} {"id": "train_047939", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["agent", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "0899d484a49a1d52fcd139ff2050981c0d475c62"} {"id": "train_047940", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "6efe83ad2deb99ec069a23136ebf6619413fba17"} {"id": "train_047941", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["coding", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "6a54dbc4aff1d8985044fdd45dbe87daf7265f41"} {"id": "train_047942", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "9b68b94d09b4c478a2c12f13af4143cecd968e28"} {"id": "train_047943", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "b12091eb2ac6ecffb5e49b8711ba0df321a2763d"} {"id": "train_047944", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "832a72ce3a851b8f2c94dfa6c5cf151d4bde00fb"} {"id": "train_047945", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["agent", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "11c82292159635777f944c7dbfe76ed12dc814ac"} {"id": "train_047946", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "C#", "moe_experts": ["agent", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "ebcc4227996926853a7f878b4554a5962ff32d76"} {"id": "train_047947", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "d5fee73425f837579556129b905409aa3691fcc0"} {"id": "train_047948", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["agent", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "61e6eb48b0ec3f7d768c010fb4db0a19d8d1f1bf"} {"id": "train_047949", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "709e5d8523ee7b2f039e2cc823afb60875eaec85"} {"id": "train_047950", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "caa1155ba3ce5b5b7bce7c66cbafda6cb6d657ec"} {"id": "train_047951", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["performance", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "a0511fd6d8206404e1f31a09427b7d4ce65e4185"} {"id": "train_047952", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "17e8212aca3f75b9d3ea8ee8b7a3de26fc3ebc21"} {"id": "train_047953", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "8a0b9e2c6c4a1d9cfb51b2f8c008be1242625128"} {"id": "train_047954", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["security", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "b599cf079314b0f278592a7e9a726cc31934ca64"} {"id": "train_047955", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "08d11247b7b7e1765324d34759090f0bd7fc0072"} {"id": "train_047956", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "1748cdc68f57df65008b8c89216704d9b0c59f70"} {"id": "train_047957", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "939101ecbdc0e4d18c819460bfe014e4a3965db1"} {"id": "train_047958", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["data", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "7adfcd1355ab7e6abb8e27f6e2da2e7356b94d65"} {"id": "train_047959", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 9,\n \"confidence\": 0.9,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["data", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "4d9ee5b179bd2999906794099041268be7ce5df1"} {"id": "train_047960", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "b3842402629c59c952a9cb7853ebb5c1980ba926"} {"id": "train_047961", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["governance", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "b8f1a53872206cb939408636b0e6a2e443aa1cca"} {"id": "train_047962", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["performance", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "7dfb39af7160f9c264a5b96ff6a6a53d09304546"} {"id": "train_047963", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["governance", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "b5911f1221a78f93d3f72a38150082e9b79ead07"} {"id": "train_047964", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "ee153d15a087da06f55037969e5d09e64a25d6c7"} {"id": "train_047965", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["agent", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "197a2637e9591db13ecf97bb90a35d26c11e242b"} {"id": "train_047966", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Rust", "moe_experts": ["research", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "ccc1be1ad3905d5422993d0cc3440a18999eeb27"} {"id": "train_047967", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["agent", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "4bdc2dc6ee8f86f7551b09822ba06cbd608f3202"} {"id": "train_047968", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "6e746b64e29a6119f113ecca88f9111f6279177d"} {"id": "train_047969", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["coding", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "1f2719a5f374643a87fbb7883d152e5e197ea513"} {"id": "train_047970", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "aef2923e575245bfcf30447ff7344708abe66888"} {"id": "train_047971", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "340ae2157df4eadb60f49136d069623ff581c042"} {"id": "train_047972", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["research", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "f28e71735e730c0883a0e41f491c6442326bb766"} {"id": "train_047973", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "688510939442cbd347716f081b94ad4e42e7fd14"} {"id": "train_047974", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "540b759600e63cdad1457c254546d40a9633b979"} {"id": "train_047975", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Java", "moe_experts": ["data", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "bc8ba07e8f2f4579a9184a08ccd0586a2d0c33a9"} {"id": "train_047976", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["research", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "92887e57729a397dc5f7f0acc594a20a9f917223"} {"id": "train_047977", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "SQL", "moe_experts": ["security", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "2757f4a2270b0439b98f01877c1d770fe1f5e89d"} {"id": "train_047978", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "04b9e68f6a3efd26ae651b9e0620850c190035c2"} {"id": "train_047979", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "302ce8822af15052766544a1983cf027d68db59b"} {"id": "train_047980", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["coding", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "aa9a962115ca06be39debd59027297661e9b9c5e"} {"id": "train_047981", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["security", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "59e22e95c38354cd24a2f9652ffa24138d2ef1c8"} {"id": "train_047982", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Go", "moe_experts": ["research", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "7d82e3bb5d2cc628f9388f140922e9d6d6a7dd5d"} {"id": "train_047983", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "5d7f06c6b34e5ec4943b640ff157fc6ce99fc77f"} {"id": "train_047984", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["data", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "d5cad8ccdd4c5ec2f7b7a67c45cd21174f19d207"} {"id": "train_047985", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["performance", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "468f34fea52194dfa73c8255f557b8bec22135e0"} {"id": "train_047986", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["performance", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "d60cf066c8c12e84170cc9c829cad5981f093688"} {"id": "train_047987", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["coding", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "848f7859bc35584c329b40229d002ac166f07c5b"} {"id": "train_047988", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "ee940308d8d8643a517d6fb4bf9fa30c65d2a276"} {"id": "train_047989", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "80936af96edef722d58553b9823258b3faf3a1e8"} {"id": "train_047990", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["performance", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "ae830ea1e6dbdf806cb7c164581a2bd0cff646b8"} {"id": "train_047991", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "ba26b2178d0d26441eb006eca2e216eb7a259f71"} {"id": "train_047992", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["data", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "51cc13f100dc88cef85d46bebe8782e42d0450da"} {"id": "train_047993", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "a2fe3cfea76edeaded90edb9768366d9e3783a5c"} {"id": "train_047994", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["performance", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "5e03356b68cbb94eb2c5c9d6bb87ea4fe081e802"} {"id": "train_047995", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "946f330286f3cb5d71b38e48c6960399f2995899"} {"id": "train_047996", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "c0a02d75a52de834bfcbfd9734d3900d64e8e234"} {"id": "train_047997", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["security", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "a203addcc62fedea8ff034152f8876dc43fb05f1"} {"id": "train_047998", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "C#", "moe_experts": ["data", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "2145c76282085b50abce191945854a9997a40340"} {"id": "train_047999", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "fa1006a21358832350ff452c33c2fe9ce554d230"} {"id": "train_048000", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "e033f01e358d017920153337a547d19548dfefdd"} {"id": "train_048001", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["evaluation", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "ee7d10fbb60764854446db025361f9a115f220db"} {"id": "train_048002", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "1c6e2245e669fcf310b4b05015c29df03d5a0141"} {"id": "train_048003", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "06966bd304fa1a508b00dc89b39a358242d2810b"} {"id": "train_048004", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "SQL", "moe_experts": ["governance", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "130088ba904ee13c3d9a1bdd9e105d70b3d65a5e"} {"id": "train_048005", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["security", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "02d0be48cc7f33c0d391bcf432c780c5ac0656dc"} {"id": "train_048006", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["agent", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "7784a86bc5bbdba62386c003add2310d7969b2c4"} {"id": "train_048007", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["security", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "a5cfa2038956f6fd1cd00167e53001048dbf54a7"} {"id": "train_048008", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["agent", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "c1079a4aa7909fa823a2275ec6369c8425ccdd6c"} {"id": "train_048009", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "59aa170e106b077416b118b92a9d59a54314522e"} {"id": "train_048010", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "d8cc09d93716a91c0df3d025cf41ac48357519bf"} {"id": "train_048011", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "4cba24632bd28a15ecdec8e7e1d3f0c52c1c5bf1"} {"id": "train_048012", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["data", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "b20522ccbcb561ec8a43639ca41963fe8e8b2508"} {"id": "train_048013", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "7b6569ff4fed008cf9dfb84f8aba4e822865921f"} {"id": "train_048014", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "f2632477376871daf22d0423d40eaae903ab5ae8"} {"id": "train_048015", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["data", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "49ad207012c3f6a922a903708b064e69abe32840"} {"id": "train_048016", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["agent", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "429626b4a4ffd06edef6779e7ab2cc272981015d"} {"id": "train_048017", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "1806ebe8f8da2aade4a9780c111d4f81f6becaf2"} {"id": "train_048018", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["agent", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "1cbf5a799a7a58db9ece25ac39ffaf46726a2387"} {"id": "train_048019", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["performance", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "cfd930c8b3bee274c30affb9344167334d839520"} {"id": "train_048020", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "SQL", "moe_experts": ["governance", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "7f93e3617c9e556ec9dc72b4ad2d69c7a70485e3"} {"id": "train_048021", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["research", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "a89313a98819af7ee0b92c8fad47c1a95d730982"} {"id": "train_048022", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "144379d37767c9b0c8b609782e359489f634b7ce"} {"id": "train_048023", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["coding", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "6317d753841faa842b3f82fe907a17c6eb48da02"} {"id": "train_048024", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "51b61e698ca85b8d2190c33cdca2b1363ee9f153"} {"id": "train_048025", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["coding", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "447326527ba2869e239cb9c6fd5c0e80431a7df0"} {"id": "train_048026", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["security", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "cf02f5cb4cc6a04ff5b7c34cb4c08aeab7174a55"} {"id": "train_048027", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "7ecd1975b6e3a11410730cc2f34283c0dfbbe1a3"} {"id": "train_048028", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["data", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "baf817791e9c11944838e7672e93034071b10ff8"} {"id": "train_048029", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "1e3e57c272156fb163e6f28e8bae65b4117a96ce"} {"id": "train_048030", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "C#", "moe_experts": ["research", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "ef7a008f6191a9dd7f951bdf38b6c32bf9461620"} {"id": "train_048031", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "9c23957daea291a5351da27c2e7a1e66efc81bd5"} {"id": "train_048032", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["evaluation", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "971f5f85d53e4b82154eb2555529930160ea613d"} {"id": "train_048033", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["research", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "43b76f2f005177bbf499164d90a4b7a6dae53f61"} {"id": "train_048034", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "352324195b5c5b2e039330728eadca4da3d50a3c"} {"id": "train_048035", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Java", "moe_experts": ["governance", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "a84b4694328591abf9342a67881e7a7bac73f686"} {"id": "train_048036", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "ded3f204c56da6af7804f67c304f65e859cc2218"} {"id": "train_048037", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["governance", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "4812198987b012fc3dd655e5dbfaa1ba216f507a"} {"id": "train_048038", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "2ee1a113172a442bc8d8c03f4dd5291ffe013388"} {"id": "train_048039", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["evaluation", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "e713336670c918157071488fa943cf466db375f5"} {"id": "train_048040", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "3bf632554d3cc7a8ab1db5054d8a0912ad163529"} {"id": "train_048041", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "0bf2373ab01ce8c36a05f6c1b08f6c9683e95665"} {"id": "train_048042", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Go", "moe_experts": ["agent", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "5e526437c3773e7d02cb829d299940f82948bd64"} {"id": "train_048043", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["performance", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "26ecf84225607c976d5caa9853e9974470fea77e"} {"id": "train_048044", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["governance", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "969d832e43a71ec58e885144ff560db45ee5fafb"} {"id": "train_048045", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "64e4f32511a48558b29080bf78561be91e78c4e7"} {"id": "train_048046", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["agent", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "06ec99375af12dca663a74182b2dfa731d7cbb4a"} {"id": "train_048047", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "b77b5fda82532730fcdd3732f0087351e0f8136d"} {"id": "train_048048", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "3c37768fd35fd837d816ff4c0be25f10911d39d6"} {"id": "train_048049", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["evaluation", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "4421186002c71102f4e7e9c20d01712d21f5f19e"} {"id": "train_048050", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "2bc3592679ad57b1ed3a63ad063961ff761696b3"} {"id": "train_048051", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Bash", "moe_experts": ["governance", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "079047e1ab7bc10ee7be630eb47cdfa4e7b59b94"} {"id": "train_048052", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "675639b678675fda494998f2d2d444da01894170"} {"id": "train_048053", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["agent", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "38d7f82a865bb1e8a44207803025a817088accc2"} {"id": "train_048054", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["security", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "926ca0d356130117ba6395d8a02949220fa435e4"} {"id": "train_048055", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "c88967ea78642e624d04dda07ab246c77c884338"} {"id": "train_048056", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "14dc1b180d4e7eb3cb7af540066b7fbd36efa916"} {"id": "train_048057", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["security", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "855bfaafef825793152389b39280e11a56ae1a80"} {"id": "train_048058", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "f652c2d9b8e36e4689da2c87f6a7386bd4d86a47"} {"id": "train_048059", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["performance", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "ccca8721c6f91e3d9cc9d8f55e2bcd36647a61f4"} {"id": "train_048060", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["evaluation", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "af4a641702e8aee3b28a5d8e2197a01e83a22d6e"} {"id": "train_048061", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["agent", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "c8e69f511f3840c5ba35ffcc1039bd20a12895b1"} {"id": "train_048062", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "773d4f13af0deb90de69dd8ac2cbf49a0eb0e920"} {"id": "train_048063", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["governance", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "feae6161c4358541d141e43b1e61c77252cb481c"} {"id": "train_048064", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "2ee1a113172a442bc8d8c03f4dd5291ffe013388"} {"id": "train_048065", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "e27dda8422c74bd772a791a9bc55581e52e038c5"} {"id": "train_048066", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "SQL", "moe_experts": ["performance", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "43703d8ad330110881c3a69e4d9cd1ee9a0faae0"} {"id": "train_048067", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["governance", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "0795dc0df3c8dbe9c212e40cf43a342a63a72188"} {"id": "train_048068", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["coding", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "a5a900a3f30a702d5c7edb210e58bd2fe1a9dc9d"} {"id": "train_048069", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["security", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "e4ce4e4974826c61c5c570d5256910d0369d1ee4"} {"id": "train_048070", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "SQL", "moe_experts": ["data", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "5da5fab315994ea277695a537dde33442bae2bb6"} {"id": "train_048071", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["security", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "4ae32b39e316f24b98bd27f47d406b348ed9ff4a"} {"id": "train_048072", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["security", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "00add1e0dcdaf19af0ecfb6b225efb6b1f8464ba"} {"id": "train_048073", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "70a8f81a0b0e0dbcda93caf89ec948c48becf549"} {"id": "train_048074", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["data", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "38e620016bf31441c2c69adbd3f1e4d2305eb111"} {"id": "train_048075", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "6b3cc378fdcec6ddbd8d2a2d20ba7e52cd9f6254"} {"id": "train_048076", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Java", "moe_experts": ["security", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "28298afb975c3a68639a86e9804c490c03905286"} {"id": "train_048077", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 9,\n \"confidence\": 0.9,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "a634b7eb96cd809556407b63f10542762675989d"} {"id": "train_048078", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["coding", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "173351ec0b8ca3cea781fdddce09fb73aba7a160"} {"id": "train_048079", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["security", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "195016aeb83247febae021274339a13b669b04c6"} {"id": "train_048080", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "472cea45bbaa6859417ce9c5662fe6724f979359"} {"id": "train_048081", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "beb03ae4be43fefa3abbee516e89a8725b73f48c"} {"id": "train_048082", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "50314f153af2fb2146e6775221f2c1e33871fbf6"} {"id": "train_048083", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["governance", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "ff16ee4a422c26b930ca1947de3229e18be4e9f4"} {"id": "train_048084", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "39afc514d566f836637ac1ea20540b47f5384109"} {"id": "train_048085", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["coding", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "f28e71735e730c0883a0e41f491c6442326bb766"} {"id": "train_048086", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "120cdbd6e062017975f9f0710b940cec3cc7df60"} {"id": "train_048087", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 9,\n \"confidence\": 0.9,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["research", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "d89bb752c2b40d4f305a925c437925521e924241"} {"id": "train_048088", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "ff6294c040e7b04a1d62f61fd35bc626bba4f892"} {"id": "train_048089", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "2d9c7b7ee5740331d6ef878a10a819d7685f94cc"} {"id": "train_048090", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["security", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "d228b8b4ce8c66f765ff7f251b1404cfdd2aa66a"} {"id": "train_048091", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "e6ce236eac5383cd280343aa2ea0bfb5eab1a45f"} {"id": "train_048092", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "23fbfc830d16d6263de8f9c05415fbcebc838534"} {"id": "train_048093", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "555ec985e96625ea062bfb44c91a3d27076c6149"} {"id": "train_048094", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "8679248adc901ecf9452fc3fb65e219618690664"} {"id": "train_048095", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["evaluation", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "2bd73d61f22d4e7041fd6be1569c1e1cf8785a5a"} {"id": "train_048096", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Java", "moe_experts": ["performance", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "acc50760a5b91e095026056edc0807c60ac85575"} {"id": "train_048097", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["governance", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "e6c88962efdc6072f3f9eaed9f710a23a968e9c2"} {"id": "train_048098", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["security", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "7e20fdb29d136349466ca4a2a19127326e9d0263"} {"id": "train_048099", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["coding", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "4825df6d9557c2f7fdc73a5f4335144f7248d451"} {"id": "train_048100", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["governance", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "711cffef9e48903408957b30dbdd2edce8a49cb3"} {"id": "train_048101", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["governance", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "f1902a4fe5738b13fed8b0b86704248af9ce007b"} {"id": "train_048102", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "69c168d9d7c26898cf777af46cee79cc85431e3f"} {"id": "train_048103", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "07426de9cc546c3f436c4f59c7d7ddcf85333c5d"} {"id": "train_048104", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["data", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "7b28fcaf294981545e2c1e0adce67f93bb5abf4d"} {"id": "train_048105", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Java", "moe_experts": ["performance", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "a89f06bfc8e73a42880f0611353d8d728f7b31d9"} {"id": "train_048106", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "1c0be31b09a5d77e88f29d286ff7e900268ec7df"} {"id": "train_048107", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["evaluation", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "0420c7b3435be9792ae43c05748ebe61ed0a70b1"} {"id": "train_048108", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "42a49b1a080770587de92d57312d896996401a5d"} {"id": "train_048109", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "73c2813380e74e014e98768a015f07cfad56f367"} {"id": "train_048110", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["performance", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "cde4a767f3e1296581f767de5af3781dc801b88b"} {"id": "train_048111", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "b553fb1a52e129f798027475ed1bb19d00fdfa4f"} {"id": "train_048112", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "de4fdb96200699cc97e92b969075d24a3dd8cd4e"} {"id": "train_048113", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "cfb0eefb79b88e321eded5c57ac95451729c748f"} {"id": "train_048114", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "C#", "moe_experts": ["agent", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "378f23da0744240cc085dd8ee1126b06e9ae97dd"} {"id": "train_048115", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "9516a6c014d8ce61b308abc65416c4c26f07b732"} {"id": "train_048116", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "d905ffec603820094dcab5392a362ab5bff097ac"} {"id": "train_048117", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "6692b5a5fdcef32aeed96e0fc4723fbbd3e4c00a"} {"id": "train_048118", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["governance", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "2e808637530f0ee002e9d278ef86d79861f252ab"} {"id": "train_048119", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "3c681c0a5743a7257bc1fd951cbdc780b05c5f3b"} {"id": "train_048120", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["evaluation", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "9ffb2aae4070fa0a622649448a3bb3d8932f66ae"} {"id": "train_048121", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "b51f34247d4d2dd78535d8ae9eaa598b2f3b5302"} {"id": "train_048122", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["evaluation", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "0b6c18537486e6d6a72f0a9309c4af2d6b3532fa"} {"id": "train_048123", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "cb33da329f879e86340edb5dca155bdc67101409"} {"id": "train_048124", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["performance", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "af1804ca6bb96fd9966ab3d28dada72984bc622b"} {"id": "train_048125", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "2d3bc534426984756cb2877672d38d255773d929"} {"id": "train_048126", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Bash", "moe_experts": ["performance", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "e71c3932d04829cb040c8ed6812b25be2b587b36"} {"id": "train_048127", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["performance", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "249d146b785e06a62c712f6ab1c669bf8023130a"} {"id": "train_048128", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "SQL", "moe_experts": ["security", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "05c1f810f922e4aafecf64c67f28c8a384ef0358"} {"id": "train_048129", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "6f06c5e28e27978c520f63b9c1fc2e6c57b7a353"} {"id": "train_048130", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["agent", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "508aed416fc204530bcd96975dadaff7b459e652"} {"id": "train_048131", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Python", "moe_experts": ["research", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "fb37479fe07da3d3a37da55892adefc122b8e8d5"} {"id": "train_048132", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["agent", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "267a4b3d689b5e29fd1444b7700b45fe355828da"} {"id": "train_048133", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["security", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "42a935a96c03f4d649d6fb7d0623a9f1d3f5b580"} {"id": "train_048134", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["performance", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "1c1156078237ddcb12f01d78d680f5574b6f969b"} {"id": "train_048135", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["security", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "f6499e3c64886a6c66ca08640d1f969265d15f26"} {"id": "train_048136", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["governance", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "631d5965abb8d840258fc5b03b07bd3558eb2f01"} {"id": "train_048137", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["governance", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "4e59b2de9c5074c4fe7269f071b5fa448fb2948f"} {"id": "train_048138", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n cost = 0\n for _ in range(budget):\n p = plan(issue)\n patch = edit(issue, p)\n ok, _ = test(patch)\n cost += 1\n if ok:\n return patch, cost\n raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["evaluation", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "35aa639c03ed1633860bafaca052a0b9b984e663"} {"id": "train_048139", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["research", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "13eb5e9c07b6b20896a37ad65067d0b6ade40579"} {"id": "train_048140", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n cost = 0\n for _ in range(budget):\n p = plan(issue)\n patch = edit(issue, p)\n ok, _ = test(patch)\n cost += 1\n if ok:\n return patch, cost\n raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["evaluation", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "eafa52dd1479dd73d483fcf474393b257cc26dc7"} {"id": "train_048141", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["agent", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "3cdc405ffc4353eda05d3e210ce235de193ffc9a"} {"id": "train_048142", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "d9d741066b2f0b6c6e9ba7bb591d4b5089526a6a"} {"id": "train_048143", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["data", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "38ba1709ce722a1bb182463851966b8e8cca2696"} {"id": "train_048144", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "0b9fd3ef1690c07aed2ed82126a12cff06862654"} {"id": "train_048145", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "4ff31c7237d2391171777f7863a0dd1ae37ca57f"} {"id": "train_048146", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["performance", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "1d40c7d897f9d0c7e4a373724e326693c23e60c5"} {"id": "train_048147", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["coding", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "7f0ab794e58fb04df248819d2c16ea6134558146"} {"id": "train_048148", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "74177464acb2f265eee3c444553554dff02a41dd"} {"id": "train_048149", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "C#", "moe_experts": ["agent", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "6c50a9daf8e3f849362d13450a5b87be6cbb59b1"} {"id": "train_048150", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "cd0057b81c844d9de517c70fb7b20743d1138e56"} {"id": "train_048151", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 9,\n \"confidence\": 0.9,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["security", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "076f414cd5eaeeb480d35c719e90c50a5c1e0392"} {"id": "train_048152", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["security", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "c34d510493cbb9a7cd32813f8ea8d1ac85e7d6ad"} {"id": "train_048153", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Rust", "moe_experts": ["agent", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "85404d24930b2d9677537dfdf61a6f5b1c6377ef"} {"id": "train_048154", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "bdad175993f9d6d8248e3e0e6034bef3d2028ee6"} {"id": "train_048155", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "SQL", "moe_experts": ["security", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "ee19a468ce50ab3e5df143eaf0c65171feb2e0fa"} {"id": "train_048156", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Python", "moe_experts": ["coding", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "87e6eb4d532ea11afecda167594fcdb0bb22eec0"} {"id": "train_048157", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["agent", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "5627330a3d3f5cfb1a2b39e9c64602012d19d019"} {"id": "train_048158", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Python", "moe_experts": ["evaluation", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "5079eaecf65cf14ea534b08ed07eda5738ea30c0"} {"id": "train_048159", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["agent", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "78fb273ea51d9a1973b9234ea374a9e80abf82fc"} {"id": "train_048160", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Bash", "moe_experts": ["research", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "a31cf1ac69ffb5d5917c3514ca2538c41bce2ebd"} {"id": "train_048161", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "486dc328f406fa576cef1d833a3537b6eee46065"} {"id": "train_048162", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "b6e13693a87aa116fd2bfddbb42235b7a6c3c1bd"} {"id": "train_048163", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Java", "moe_experts": ["data", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "8121877b2b746c80f2605d3057b299ea0c8850dd"} {"id": "train_048164", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["research", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "2b78ae7e41d080b2f0357f6da73b1dc9f80fc7dc"} {"id": "train_048165", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 9,\n \"confidence\": 0.9,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["data", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "966a3a825c2ee3a0fa401ba869dea3bb80760416"} {"id": "train_048166", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["performance", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "32e29416f1c5c2baef4855920fb22698d16477d0"} {"id": "train_048167", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["security", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "7c47594b69982e70516ccae274b887d3bc5c6c19"} {"id": "train_048168", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["coding", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "85af969ccbc0ab2a6b0381dc9d7f244f56af1e39"} {"id": "train_048169", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "5d69b3865e039704c5491ee9a4680bdd73da8295"} {"id": "train_048170", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["research", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "886aec2d53f980d0954137445c6e6ceec2299c38"} {"id": "train_048171", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "540ed2ca75f83d0a6abb7e0b6c9d747301eda6fc"} {"id": "train_048172", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["agent", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "eab2966672a10614e4da90b5e82f02ae2d3d4279"} {"id": "train_048173", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["evaluation", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "6092639562db6e52469c274b7771db373139d73b"} {"id": "train_048174", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["coding", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "b4cf71fa0c0edc79c02e395e1ffe8c9750495c01"} {"id": "train_048175", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["performance", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "3924203cdad3819b6b6f651c5dd61a3c1048479e"} {"id": "train_048176", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "a599e6dc0f173b20d0c9ab14aee723faa3d33e5f"} {"id": "train_048177", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "493888bb07e89a38affd775d519a9d0c67e82bc8"} {"id": "train_048178", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "1880d25121c7a086c98043276fac782ca0fe48e7"} {"id": "train_048179", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["governance", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "ab47561dd24e7510d2708fb4a150f2f8ec3e9719"} {"id": "train_048180", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["performance", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "420c3c60622796f5843b76b5fb18d0d0f925ced3"} {"id": "train_048181", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "210a2ed9b844bdf8e99500cf402b7aff90684513"} {"id": "train_048182", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "f348728ba405707257d1232e8927ea7594c64c6e"} {"id": "train_048183", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 9,\n \"confidence\": 0.9,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["security", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "b9be807f48b59524572a939540f1a001bddacccc"} {"id": "train_048184", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "d5e56b4daddcd53415f83f61fbfb6126fcc5e5b8"} {"id": "train_048185", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["performance", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "7130dc4d7d51c9bad96c4b15a84c19a2f0337eb6"} {"id": "train_048186", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["security", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "6b0d4829439a7303d2a1d0756386388c656dd16c"} {"id": "train_048187", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "6a0239bc8b675fdc37d8022d1a8aa2dd25426267"} {"id": "train_048188", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["security", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "4d1a2b6648752d621114ccd4e5e45646053468aa"} {"id": "train_048189", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "d3b17a19b54270126d35310868ece6e5e5c64f4c"} {"id": "train_048190", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["coding", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "b57b626e0ffa79d05143208acb65571dc6fafff8"} {"id": "train_048191", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "f886b1402c6f32d1802f11114c9a97702ff41af4"} {"id": "train_048192", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "11ceab9fef1b2e416e164b3ee700b9569d425ec4"} {"id": "train_048193", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "2ae6a3d2e9b285236b58591d45c310e9acb54879"} {"id": "train_048194", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "a7e608f36ec16775b5d642df0b37eaaf9fcd1756"} {"id": "train_048195", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["performance", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "05ff15e1c11f96a0b5ca9e6c1f66790a01fe3004"} {"id": "train_048196", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["governance", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "8874e8cc5edd84f5d52d222da4ab55f3448901e1"} {"id": "train_048197", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Java", "moe_experts": ["data", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "ec75846631d3d3e6713204bd036584c28f0a715a"} {"id": "train_048198", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["coding", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "82a4265a8dcbe6ec861bcbf0fc13b23e3efc38bb"} {"id": "train_048199", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "39fd641b58200730740c649fd6e617f9b3272315"} {"id": "train_048200", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["coding", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "67ce30096e2c9dcbf941e8cfb3e78f00c7817fd7"} {"id": "train_048201", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["research", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "ed4cf7da4780ba1390b461af2a7a00ef72646d19"} {"id": "train_048202", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 9,\n \"confidence\": 0.9,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["research", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "5d74221b8a971d8da055f14ae7b15d4acb284b03"} {"id": "train_048203", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "80bd1042c864d179a88f6401ea0dc8345f3d409e"} {"id": "train_048204", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n cost = 0\n for _ in range(budget):\n p = plan(issue)\n patch = edit(issue, p)\n ok, _ = test(patch)\n cost += 1\n if ok:\n return patch, cost\n raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["security", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "41641cdee2add8de7e03c7a2635e00382a05f05b"} {"id": "train_048205", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "d4a96fe9b47331240f0d17415995dbbec19ae8a6"} {"id": "train_048206", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["data", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "c5a41a318874bf32c56e40d9d3de55bcc689fab3"} {"id": "train_048207", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["security", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "3ffdee0eed38dea4161d13ee0b53d922654675a9"} {"id": "train_048208", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "20570622e50bc71c601e5727cecdf844fcd8d7f0"} {"id": "train_048209", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "7d58a0e5dedc8fd5304906c2834487222a7c941c"} {"id": "train_048210", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "a7e40c678eb26e0027006b1aca7b761c0eb5d073"} {"id": "train_048211", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["governance", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "ff6bf48bb8c6a63444d0005e1823ccc599aff13e"} {"id": "train_048212", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "f7584da0095f0392128a77162ba12b898b9ed351"} {"id": "train_048213", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "54a39e664aa04b47595dde6b02947bb5d9bb3e39"} {"id": "train_048214", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["performance", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "0a6d26845fc9f5a80e7b7e75fb8c73af81f85423"} {"id": "train_048215", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "f507938f1a00e277a79cae7aa448bfa3f92f2ae6"} {"id": "train_048216", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "SQL", "moe_experts": ["governance", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "98e4bc2961ebcfc87b9fbe391db3d7c8496f0404"} {"id": "train_048217", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["performance", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "a2347f64cd70cd4978a906cb81eda68e32f36891"} {"id": "train_048218", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "4bdfef78a48acbed494bcb74ded716bdf7bdd9cb"} {"id": "train_048219", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Go", "moe_experts": ["governance", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "2a0a8dc4c9da431dfc9c07c1c3bfb1505e56176c"} {"id": "train_048220", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["data", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "a4cbc0060a589adaec3c5201c3377990ba2dfbaf"} {"id": "train_048221", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "76f0cf2609247867d002609af2b4ea8591968d3c"} {"id": "train_048222", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["agent", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "43312b87109dabb6789c0dd1bc857a5c0be51340"} {"id": "train_048223", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "ab3434b2981e0de2ca7ebe1cc1518cfd485898c0"} {"id": "train_048224", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["coding", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "5014c3f8db8c9341bf573643c116d263843529cd"} {"id": "train_048225", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["agent", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "147421f8723ce46fe230158a9501fe8bcf960db4"} {"id": "train_048226", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "cff88fe6be4de375e8be7239967106fa61966f41"} {"id": "train_048227", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["agent", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "ab7675e78433d40cd958d204737f44173ce385b1"} {"id": "train_048228", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["evaluation", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "6a028cc6a11c6d265cc9e7880cbec96c69173683"} {"id": "train_048229", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["performance", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "c16b7d841d55f6708e20384e96d4d871bc67abe1"} {"id": "train_048230", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "f9a079539a0cba336b7f02db44c3683ba4933965"} {"id": "train_048231", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["coding", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "ceace82ace8f34fcc6b59ed337b4cc41de2e441f"} {"id": "train_048232", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["performance", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "d4224f7210d9467d5fcfb806a7c307ff929bcd7c"} {"id": "train_048233", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["research", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "1b82d2a85e9dd3efa74cc7adfd752dab68dd2296"} {"id": "train_048234", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "edebe45e98f68dbae8031c13e94521a3ff0adb99"} {"id": "train_048235", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["data", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "35d1f44082a29837b8754034980c12361208af6b"} {"id": "train_048236", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["coding", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "d61bcbcd5ffa525b522e12f8c0172292c85c1287"} {"id": "train_048237", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "15e689c3d3d8672132739ded0e40130e1bb19d55"} {"id": "train_048238", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "9bf4d79f94061e4e784061bd386d298ec2e80802"} {"id": "train_048239", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["governance", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "75106ed73046bde58fdb5e5cfbdebffa2211ba91"} {"id": "train_048240", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "ac146e156f21f9d96b16fdb1d13828ee3b5fb2c4"} {"id": "train_048241", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["governance", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "948f7d78c2c6542cb76bef95aa664908119227bc"} {"id": "train_048242", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "C#", "moe_experts": ["coding", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "b54f110924dd1b0abf3266ae5b3996a5d7cb82f5"} {"id": "train_048243", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "0738c73987c86b492864f8464941401e505add5a"} {"id": "train_048244", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Bash", "moe_experts": ["agent", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "e7b7526900f5696eeecc0a3dfe13e28272204e8b"} {"id": "train_048245", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "7d13043aa55abfc83f4ad14f2e70f3deab074321"} {"id": "train_048246", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "5a93f66acbc8e4e78d2e8d545fdd2b80eac0350a"} {"id": "train_048247", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["evaluation", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "1599889f21310b4fe8434041a4babf9a72238395"} {"id": "train_048248", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "8c5df1ed2612b2a755173d3572799e01c7f18753"} {"id": "train_048249", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "ce07c98a5d9e69f65a5f54c4468647bf066cef67"} {"id": "train_048250", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "2debf7ca372b40695213ba61900eeca7fb6fccd7"} {"id": "train_048251", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["security", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "ab3434b2981e0de2ca7ebe1cc1518cfd485898c0"} {"id": "train_048252", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["evaluation", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "5c4327c7fc2a811a6ace27580e4df3f6a653fea1"} {"id": "train_048253", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Python", "moe_experts": ["performance", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "d60bc1ea948006d53dbfba99ed1d7ac5002c02a7"} {"id": "train_048254", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Java", "moe_experts": ["coding", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "90abe3c34e80ff67d2dd27be164e5b1fb29dc12b"} {"id": "train_048255", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "8085cdf29d09e2a833ad99dbfbde8ffd5e431be9"} {"id": "train_048256", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["agent", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "4f086522f1d7c9c9ec1bd9270861c8426a31b3ca"} {"id": "train_048257", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "182b6b0ea544b4ef3ed9afbecf916c3aa8b0d37f"} {"id": "train_048258", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "aec9f6411775855da5eb1fe61eb7ba9bb9d7dc45"} {"id": "train_048259", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "e81cb2917754a948d725b58b6cfe1da31885a4dc"} {"id": "train_048260", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "2f60fd62d38e7e9da1770770afbaa4f0f314d1fa"} {"id": "train_048261", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Python", "moe_experts": ["research", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "cd4b57f670160adffa95d76f2d1a97f63c172790"} {"id": "train_048262", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "efa3a9a020da16f72f5f7a31df20d676c8c63f29"} {"id": "train_048263", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["governance", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "7c1161c7f59118faa01a5e5957fbe2bc88939640"} {"id": "train_048264", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["coding", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "84443dcc294ce31d9eed74f5b488a398bfe99f5e"} {"id": "train_048265", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["research", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "5732f75af6592d5a3cdc49a533cdbaf21ae20ee2"} {"id": "train_048266", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["governance", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "5930b51a4c910164becdc40eb9b639c98ca386c1"} {"id": "train_048267", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "dcbb4246032b474761d8961c44f8e89aba8ff304"} {"id": "train_048268", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "cbbcec6b0cf9cda709a3d73f2fc78424b788fd5f"} {"id": "train_048269", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["data", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "836f2a71bfe4d59eb391e1b34fe66d4d85d502cf"} {"id": "train_048270", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["evaluation", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "201aafa87f4ddb1a346fda92d4737180a2e15ca8"} {"id": "train_048271", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Rust", "moe_experts": ["performance", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "e0d745b68d2eb87331caf60e1742c801032392f2"} {"id": "train_048272", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "cd1db49fa51086f52e368e71278d7717d9eff71e"} {"id": "train_048273", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "38dc2be3261dc25d9b171b228ea452546fe4cd5a"} {"id": "train_048274", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "5cc86bb599de4afe9cf6bd5599fb27766bda4be4"} {"id": "train_048275", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "72cb5844a05ac23b91c6e6a216fa473f42c6b029"} {"id": "train_048276", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "f785e2a1f58bff737d815a1380f492c651a6c1c3"} {"id": "train_048277", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["research", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "7df38f78c8bf6e90ae85375554fcc359659f18ca"} {"id": "train_048278", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "1540d558ffcd7a1b6df058864ec75a7c354f7d14"} {"id": "train_048279", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "C#", "moe_experts": ["coding", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "c1f78857508df826592158d0e9553f6ea66f9d67"} {"id": "train_048280", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["coding", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "8088ad2c6ac9daea741438c4915b042b79b212b9"} {"id": "train_048281", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "1f73e11930b66938992ff0c12d2abdf6cca9d5cf"} {"id": "train_048282", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["security", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "75d397f4d0c1a786c8cf120b52382f29ed7f205b"} {"id": "train_048283", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "e53232e275d0bce16a5602983a1bbbfa8df4bd6c"} {"id": "train_048284", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["data", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "96af6c30ad6920453d5aa52ae52d9d17618f4b50"} {"id": "train_048285", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "9ca245306e3b4f99b42ce81da630672132d8d4f0"} {"id": "train_048286", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "b0f14b6e73b3f7fdd33d93e9c5442f0eb48b3692"} {"id": "train_048287", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "2090eb414cfdb872f3103ca98d44e60135811724"} {"id": "train_048288", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "61e78ef10eae03ec33575f117bbd3a46c563f3ff"} {"id": "train_048289", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "2d04130927f1ab2d4321a481f3b9ed47365610b6"} {"id": "train_048290", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["security", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "dd50c5245af94ba7a2d6e0ea62ae58137d1bc12a"} {"id": "train_048291", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "b8c7520e5929ca6d67e6aacbf7140d853d0ccdf3"} {"id": "train_048292", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "f2a38448954b3eab613a2fd5d4d9fd24481912fe"} {"id": "train_048293", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["data", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "ec509c2d0a2f4eb91157fe5f23aadf996844c0f9"} {"id": "train_048294", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "e2945353561f1337c3080bb7ee12b6d4620216d7"} {"id": "train_048295", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["governance", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "60415ff09ddb0ca01c392919f85db34f0eaacb79"} {"id": "train_048296", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["security", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "351185c7bb13413282ff375936b3a1ea00b8abca"} {"id": "train_048297", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["coding", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "0b7a2ce5f783cd4fad66dc543ce6a043c1fd78ee"} {"id": "train_048298", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["coding", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "790bd047af69816a4081b718fcf71ea55ae24621"} {"id": "train_048299", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "fb24dbdcb44a0e24f2abc7e3089a2fa585096ff5"} {"id": "train_048300", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "b79eacdca6f93cbff44bf33fbf45bbb7ebf7c183"} {"id": "train_048301", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["agent", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "8b34b2477481106da00d7cb55dcacaa7a0502b08"} {"id": "train_048302", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["agent", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "aa0869b4242b3fcb6ef2885ae3ae06c5dec1e234"} {"id": "train_048303", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "915afac4832676a45d9ba2394b15a18254252183"} {"id": "train_048304", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["agent", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "3f2daaa9da64bb474faf682efd64efbe3447fa88"} {"id": "train_048305", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 9,\n \"confidence\": 0.9,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "baf81c7a6e8008d5bd9f9b9973d2c9cd5f730a59"} {"id": "train_048306", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["coding", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "62caa9e90e28e776a3ec84f065aa7d54999bd916"} {"id": "train_048307", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "6b1018b4be3ff30d1af5b70dd896a2e87d021b94"} {"id": "train_048308", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["security", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "ccba1212b00654d7230cbf089603bb64256079c5"} {"id": "train_048309", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["governance", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "238124a482ce10f0a35bbdfcf6d80be04b679ab8"} {"id": "train_048310", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["performance", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "601b62fcb6e6cf3900b47dddf3437cd9def0deb9"} {"id": "train_048311", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Bash", "moe_experts": ["performance", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "a31cf1ac69ffb5d5917c3514ca2538c41bce2ebd"} {"id": "train_048312", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "7ab3315035a7d411958e94c59179431b8f1e39ee"} {"id": "train_048313", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["agent", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "ff6294c040e7b04a1d62f61fd35bc626bba4f892"} {"id": "train_048314", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["governance", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "1f61b6b6c6bd59df7d35f0680432ad3b0d9e36b9"} {"id": "train_048315", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "13d8f705af9bfabc4ba6af31c6479e2f57a33f97"} {"id": "train_048316", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "84d7dea94328f2ae7bb1276d7a32ebdada0e2e74"} {"id": "train_048317", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["evaluation", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "02740d1df984a6decbef5f1adda62039dc6a5467"} {"id": "train_048318", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "f9b3c8c948782f29d9d75e685fdfb3b4f50d64f5"} {"id": "train_048319", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["governance", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "00dbd27d1126acb24fe7742d90bcc3ac509282e4"} {"id": "train_048320", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["coding", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "62db254e476330b77d45cb502f82cf1b93aaa721"} {"id": "train_048321", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["data", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "06722de3f9f42c0f332f41277bf333d80c5e92f2"} {"id": "train_048322", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "6e6f4f6d7a6fed86e9f9d58dbd086efa369912a0"} {"id": "train_048323", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "a1af557f126cbd9f1f112183d30459699ab85479"} {"id": "train_048324", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "24acc5213ca1e9aba50a5cf09a5a30267de9c416"} {"id": "train_048325", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["coding", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "de363579a6f4c64e956e6baaafeaf8acc095ba8f"} {"id": "train_048326", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "581141e376c77cffd8c2bd0b6758d263d30cd884"} {"id": "train_048327", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "647719b53f3e671b493997947771d92912489a6e"} {"id": "train_048328", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["agent", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "0b5f078931cb0220c1d45cefb0a671ef73aa0e1f"} {"id": "train_048329", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Bash", "moe_experts": ["agent", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "a35454f4bda5a0dbd7ef9d4ad1d05ecdbd21e591"} {"id": "train_048330", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["research", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "c4e8b8482179e775381d9108c21be514e63f4e1a"} {"id": "train_048331", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["governance", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "40db7c8fa134038fb54f6db921ff227d7387c388"} {"id": "train_048332", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["governance", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "fc50f527cd983137614dc677692d954e1550d54d"} {"id": "train_048333", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "1974894efa2ce95e5f91c7dfb020e65cea5fc2c8"} {"id": "train_048334", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "fcb5f489e544ddb841825419509960da8cf327ce"} {"id": "train_048335", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["research", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "d9c29f58a07b3aaabf5ccaa7ce40058d18003c7b"} {"id": "train_048336", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "68fc7531bace5be46990d38bf062b9e2421a4883"} {"id": "train_048337", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["data", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "ef2ae055623054c7f9ab2590337ce91a075f75ac"} {"id": "train_048338", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["governance", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "39ade47c9ab33873eb8b433c357df15e748a7e68"} {"id": "train_048339", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Bash", "moe_experts": ["performance", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "f46ad2ee9d662cbf607db6436e6dc8a827e87216"} {"id": "train_048340", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["coding", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "88797202c08dbdd5d39e673b8e10511d86e557c0"} {"id": "train_048341", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "62381e2beb8173b4c35166b8483bde1aea22ed1f"} {"id": "train_048342", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["research", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "6769d0be7123c4b4dc3d2839064eea306e4a0826"} {"id": "train_048343", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["agent", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "1f869aeb0629686554eb2b87d0f38d04af6d2f0b"} {"id": "train_048344", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["governance", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "6dfaeff1716776bf0aa7caa8035d203845174869"} {"id": "train_048345", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["data", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "dfa63af78b65c3868b08ae75f8c7fffa74be5da7"} {"id": "train_048346", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "72682fb9057eb9cd6615c465502eeb613810710c"} {"id": "train_048347", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["governance", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "f7a487849651b1bffb21b1d725a12f1229d842df"} {"id": "train_048348", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["research", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "28f51f91ba36985c2e833dc9b8e20bd0f6083f2b"} {"id": "train_048349", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "c5fa512ff6fb50f1cad46f9483c03c7963991ac8"} {"id": "train_048350", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "c8339fc98bce41c75e99a68d65d54b8898a01823"} {"id": "train_048351", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "f77b09e98021c51d2c1a6430564a1a7be249cc08"} {"id": "train_048352", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["coding", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "9d675e65bf253363f186a93d588272a404e37fbb"} {"id": "train_048353", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["agent", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "29aa510ef40a10ffdc91f65faf20dad16e6b8f5c"} {"id": "train_048354", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["research", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "77c06836e6bed26bf6b3fc3e881b5f2cfccedb01"} {"id": "train_048355", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["security", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "7adfcd1355ab7e6abb8e27f6e2da2e7356b94d65"} {"id": "train_048356", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "ce19da4006541de060751c9c6d013ee2130ef51a"} {"id": "train_048357", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "3f5fda589bf383a7942d0be4101bd72db4f84326"} {"id": "train_048358", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n cost = 0\n for _ in range(budget):\n p = plan(issue)\n patch = edit(issue, p)\n ok, _ = test(patch)\n cost += 1\n if ok:\n return patch, cost\n raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["security", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "a0d2931dff0d3ed68785ca9c4ce5e2ee3c5f6daa"} {"id": "train_048359", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["agent", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "16fa9e082773410e14eba211fc23f684d9d55f59"} {"id": "train_048360", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["governance", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "b2d79f1601d08ba455c0d76d285a90c904e8ae4e"} {"id": "train_048361", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "76d08805dfc19d06439e6debd321fb842ac7101c"} {"id": "train_048362", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "1363e2f445498492b63a4411b82ee7ba0d5c5d0d"} {"id": "train_048363", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["performance", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "4c7788a0c640865d685bf3d768af397d695c27f9"} {"id": "train_048364", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "2d4530b2cf77931cf3a4729a7708c998f2014fc0"} {"id": "train_048365", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "bd832b72be7a1f956430959e12430a818425e003"} {"id": "train_048366", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "b159297e4a0285be01bd9f7ca20aca620399662a"} {"id": "train_048367", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["governance", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "f067571dce211d135a511f1298e6e924472acc8f"} {"id": "train_048368", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "3ea78ef44d7aeb27ccfd5f0e2c1a4ca6b5623c4c"} {"id": "train_048369", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Rust", "moe_experts": ["performance", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "cb35ae136a998d4e40962d3c94e729c35a59b2a4"} {"id": "train_048370", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["security", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "f0115bb715da4b002da0683122ee9f8d0560ca41"} {"id": "train_048371", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "1ce29d25fbe68c3a54d9176722159ad593d51367"} {"id": "train_048372", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["agent", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "665ed4b4e0b8742b231de0407160a2adbf2d4600"} {"id": "train_048373", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "d43387e6ef70a660f45a0cc449cff579c035f969"} {"id": "train_048374", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "fd318346bcb3b193229cc16f19b50d65386f4c06"} {"id": "train_048375", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["security", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "bdb7dd1f8f153183a4e3ea3646b09242691cd764"} {"id": "train_048376", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "321072ec108a0475034a422862d90f9afd4d5559"} {"id": "train_048377", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "6f6b84b5285df474e47d01a1b66b623ebf43ec05"} {"id": "train_048378", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "bb92290ca82a6758551c470365f74ff17771d818"} {"id": "train_048379", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 9,\n \"confidence\": 0.9,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["research", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "efa8e6bafac834f126a166d2ea30cd4599a7afb1"} {"id": "train_048380", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "389ff5de8cd88a981994da186c3eec2fde04f9e3"} {"id": "train_048381", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "92677ead46be9f14ba52e56f68e990a6d59662af"} {"id": "train_048382", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["agent", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "354083d148cc9dd93976843f62ef5d65f586acd2"} {"id": "train_048383", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["research", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "645592ced75782e6fc2dfe2b0435ea8f4fc52761"} {"id": "train_048384", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "fb2e7132bb0e59090c5558108d29ca479a290313"} {"id": "train_048385", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["research", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "baf817791e9c11944838e7672e93034071b10ff8"} {"id": "train_048386", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Go", "moe_experts": ["research", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "7250b414b8f3465e22abf3ca15b48a612c06d4ba"} {"id": "train_048387", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "b4d46133b846c2125626c0d0057ee890d65233fa"} {"id": "train_048388", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "2f0425f1bd64ab2dbdb3d0364eafec8a59a1f223"} {"id": "train_048389", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["performance", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "26f04477eaecbe56afea7ac2a7bc663653633478"} {"id": "train_048390", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["evaluation", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "bdcdb3ef8501ae959081c882b72c56c8b8512be8"} {"id": "train_048391", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "c96ba243a8852f67fef5d2d6d74d8d9b6d025f20"} {"id": "train_048392", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "1a8edc7d17cf28b88e8d035805ea9b0959484a49"} {"id": "train_048393", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["governance", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "506a757c410b90fd6180a94b98e6e13b7fa1896d"} {"id": "train_048394", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "1f588795e4a55da4588e2e6a4cae372c02ea1194"} {"id": "train_048395", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "e30cb27e2e427064f82c0b1bc70ed665eea4f225"} {"id": "train_048396", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "a8e9f5ee8f57e09b6cd3b3135d6e0c0d41b895fd"} {"id": "train_048397", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["research", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "40e34f2328444582d1824883fd9dbf67b08ace6d"} {"id": "train_048398", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "SQL", "moe_experts": ["security", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "65d666c730e3920275dacb5eed4aee35428ff585"} {"id": "train_048399", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "38c547184f396e1219360e5e04e16122a1578f54"} {"id": "train_048400", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "ee783e2f5d81cb7e49c9b527291b343135c6b3e9"} {"id": "train_048401", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "af3ec344e09530ae956ab75d7298a085447126a7"} {"id": "train_048402", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Go", "moe_experts": ["governance", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "0f1f10d7aed1cfeb05050e93b4833eda5557e297"} {"id": "train_048403", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "2a23d3003ba37c30d60ee3936bce435573cf7e70"} {"id": "train_048404", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "e69bab067e625a46d4106639e6b2a1a281b0915a"} {"id": "train_048405", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "1e4b7b3851f1d06a5095148cb174faa62f35177c"} {"id": "train_048406", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Python", "moe_experts": ["research", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "24e353e0161516af42f593435b66851382b5e6a4"} {"id": "train_048407", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["performance", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "2325faf740603a82df066d181fb1a314308c662e"} {"id": "train_048408", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["performance", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "78e53ddaa79a2a96fa9cd12b712e34b617f9273e"} {"id": "train_048409", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "52d50722dc906c197ddfae27efc96bd889e1b7e0"} {"id": "train_048410", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["evaluation", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "3247d3499a743878191221f1b07eac265618f498"} {"id": "train_048411", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Python", "moe_experts": ["governance", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "399737f25d00b0069931f06baebc0f742a9ae033"} {"id": "train_048412", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["security", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "1f2032437b3c5718e59a94ad8edf60ec0a723f36"} {"id": "train_048413", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "d8b83929b9a1edf7203c9bbec349a86a25183d09"} {"id": "train_048414", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["governance", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "84b862e3d9e9d0b53641e3f4d3138ab9f366f245"} {"id": "train_048415", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["security", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "86056c4caa7412abb7246293b985975235be18cf"} {"id": "train_048416", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["research", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "eadc287ca81c5c0962fe214cbb2a42412c16e64d"} {"id": "train_048417", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "e7a3b67119c0370a9bc0842ec1ed98b280956c47"} {"id": "train_048418", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["data", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "373c027f46b692d7aaf40efe0f776835efb360ae"} {"id": "train_048419", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["security", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "14e0c89382f8b94ac4e3af5b4f12da8d6d6923d9"} {"id": "train_048420", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "9db78bc4c11f42a7694f59001fc49dfca0868159"} {"id": "train_048421", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["data", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "7789271c12dc225b2fff8567e59c796b6626e1c7"} {"id": "train_048422", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "da24c9cc83b6987dfa9583679c63b8652852a81f"} {"id": "train_048423", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "db155495dce08f127ccb3b31c36f36a06809b821"} {"id": "train_048424", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["coding", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "b391c651ec9d5e9580d454005a6cc3fad2731a7a"} {"id": "train_048425", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Go", "moe_experts": ["agent", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "02c35d75a58f6b5e1f4a530da7c7325ef4bdd995"} {"id": "train_048426", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["data", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "556537289def2c613948f30d52810e6deab170aa"} {"id": "train_048427", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["security", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "53729681b1668f729ba136dcbfa57b9c3fb0c1b4"} {"id": "train_048428", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["research", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "dd598adde0538a0bbfd43332c25f5362eab347b5"} {"id": "train_048429", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["coding", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "d892ece990ada534539efe8c97492b768b8eb1a8"} {"id": "train_048430", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["agent", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "3ba190554088cc072d43cceb2f3d7748f747fe88"} {"id": "train_048431", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "12a2cae999e41ad0f183afaef8279bfb98295126"} {"id": "train_048432", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["evaluation", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "e2430f05c3b2712158dda100db24743c0b72ece2"} {"id": "train_048433", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "a7698f4bfd4a94a52969828232d4c28b29bccfad"} {"id": "train_048434", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "5795d7df9ac6c8fb5b4bf8e51d86fef7c90eb5c7"} {"id": "train_048435", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["data", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "a0b8f54e6eda22f72d909e523b9a110e9e1abb36"} {"id": "train_048436", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "831b6662ca9e2d9f020bf48d8c14dd66a78c4ed5"} {"id": "train_048437", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["governance", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "61c6b062f4cc89b7175eeb7bcb6922e2fb421563"} {"id": "train_048438", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "ac018b34fdce58c50671e2014e9fbeb0cbb4c364"} {"id": "train_048439", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "d1515646625631d4e7f4c82bb3caf09b937f6330"} {"id": "train_048440", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "483319a9808d023927b09d94bb9ede9f78961492"} {"id": "train_048441", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "b85f67221aaee8f5e4fa9be853851e2dfcb00632"} {"id": "train_048442", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "72f6594a28c43f18247e1b46bc2736b000eb76a9"} {"id": "train_048443", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["data", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "ba1cfeb911de521e726a097ec22cc1784d46f916"} {"id": "train_048444", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["coding", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "53ede357c46e655b971be71ca83cde3dbfe71d3c"} {"id": "train_048445", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["coding", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "9c781d0850d7e6eff9c47eaf1ccd5a90c379812d"} {"id": "train_048446", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "18be66614870ed50ae3567c0d54f71f287a650ee"} {"id": "train_048447", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "SQL", "moe_experts": ["security", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "afb37d24c8a6fc99a9beded795481bb39e63cae2"} {"id": "train_048448", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "b4f191ceb1d35fe0f3ef87902b7ec178f9d7a667"} {"id": "train_048449", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["research", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "8568ffa69749b6adaa2490970d233c80566eb883"} {"id": "train_048450", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "f338aa46d8a81ef9032bc14a792395847a85520c"} {"id": "train_048451", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["data", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "17070f88b22c2df29d7ff514370a499bf435ebcd"} {"id": "train_048452", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["research", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "4e75d8a1027f37af352fa56204cbaeeb81a98393"} {"id": "train_048453", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "d5ae096e703c4681511e872f7303b3e915e70740"} {"id": "train_048454", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Bash", "moe_experts": ["data", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "ef1d91d115c2a1377419cb1cb6a9105c676850e9"} {"id": "train_048455", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["agent", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "71fab7421dc9461c4323c3cc5de6f6da39cad949"} {"id": "train_048456", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["research", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "83c9200aaebdedf9e1c54639a748740741b0fb5e"} {"id": "train_048457", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["governance", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "354083d148cc9dd93976843f62ef5d65f586acd2"} {"id": "train_048458", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["research", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "cf263be34b49e68bc6312d884bde6960eaede738"} {"id": "train_048459", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Java", "moe_experts": ["agent", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "aa6a287da58ba959ae21d9ec9bd9bc03cbbdc76f"} {"id": "train_048460", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Python", "moe_experts": ["evaluation", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "1bbeaf73337885b85907d62fef6d958fa11ed08a"} {"id": "train_048461", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["data", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "73a6f57777c73502af9461604f3a80096e0a442f"} {"id": "train_048462", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "SQL", "moe_experts": ["data", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "1bd3ea4e8192ad88e074d60e7c3beaf0fd3ce347"} {"id": "train_048463", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["governance", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "9351c26c26f111af306e76ed7ce63b4adb668114"} {"id": "train_048464", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 9,\n \"confidence\": 0.9,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["evaluation", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "cd4a8bec604f70bcc35ba06f02bbff6162e4635e"} {"id": "train_048465", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "4f030bc84dcb0443bdf8be15647eb8171bd32562"} {"id": "train_048466", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["performance", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "bd46bc4f68fb995eaec63cb832307b236b01485d"} {"id": "train_048467", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Java", "moe_experts": ["evaluation", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "2ab2b6c5967c4f7a78a787b507af29093f2bd0fc"} {"id": "train_048468", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["research", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "aa71b5ef3d35c6239e3414cf6a1044fb10011754"} {"id": "train_048469", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["research", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "3477ea813843733491046f8a82f4366a2e6740e0"} {"id": "train_048470", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["data", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "9923d466a242fd4d857d8849d2a4d12e0e10202e"} {"id": "train_048471", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "acebd3f42d0b18c9ddb636ebf0e1329b4548d575"} {"id": "train_048472", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "5aed8cea893eac75d0add5c2dd3b7e0b802db8a8"} {"id": "train_048473", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "4375179129c4af588535bbb925676cbee4f9ec69"} {"id": "train_048474", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "841b1e1de15ab0d96222a22ac2cd73a47725008e"} {"id": "train_048475", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "214fe11f103cc850fad353c3ab4cf903e85fb857"} {"id": "train_048476", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n cost = 0\n for _ in range(budget):\n p = plan(issue)\n patch = edit(issue, p)\n ok, _ = test(patch)\n cost += 1\n if ok:\n return patch, cost\n raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["governance", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "7c7956b27f64d4a5ce1bf680ac99232c750194d7"} {"id": "train_048477", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["performance", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "79a81aa55b040864ce80625bd33f65a1c9f48091"} {"id": "train_048478", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["security", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "0146d43e27a02c79f6669d881a73462fe84072f3"} {"id": "train_048479", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["governance", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "13dd2f11f0eb279c0ec5952240397b5ce88c4aec"} {"id": "train_048480", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["coding", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "3970a08c04fd2604f341ecba23ef64c5de4698a2"} {"id": "train_048481", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "f85ab4bc20f5846254a675a3c00aae0330185b6a"} {"id": "train_048482", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "2bd66369e420db151d0ff41493dfeb0e82cc7e03"} {"id": "train_048483", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Python", "moe_experts": ["agent", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "9f0677bcb554a76ce8bd8c03421b653036a6f7c9"} {"id": "train_048484", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["agent", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "d8623bd4303c33853db78c2482be59effe9f255a"} {"id": "train_048485", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "af850f4262e7495fb208e62c63c7930c2523b8f0"} {"id": "train_048486", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Go", "moe_experts": ["performance", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "1f368e8c550742776f487286f177c7c2faed8397"} {"id": "train_048487", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["research", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "00c4f5938098bc7a2a6a73624efc6ca0b0467713"} {"id": "train_048488", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "e368e8fd10e3b180abfdc2e17e490f1096aa8533"} {"id": "train_048489", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 9,\n \"confidence\": 0.9,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "2a76ccf06d484daef54aa2fc3127ef6d362c363e"} {"id": "train_048490", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["security", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "a314c894c54fd4927c7bda72ed18fb9bc4619237"} {"id": "train_048491", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "91e738f7a37277455dfb587d3358d8f4800f877c"} {"id": "train_048492", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "e908c6737c50b36b69511bc5a2aba243a89002c3"} {"id": "train_048493", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "73cd8ba1321b69c7013d115fa7470aad3d2ba186"} {"id": "train_048494", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "48b76ab28adb8c067df9be66090530c79651938b"} {"id": "train_048495", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "e9b39cd0c7dedaecd8960273646747197d086c5d"} {"id": "train_048496", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "004b38693c97d4ce7477b0221c65fa25d7c37b1a"} {"id": "train_048497", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["performance", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "2da4fc709ca7001eb874c74f449bdd0a56b07bcb"} {"id": "train_048498", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "cef669ed1e326fb72d3d7b292fd5e3be89f57570"} {"id": "train_048499", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "1cd28719665934b55db7b9c835cb3196529dacda"} {"id": "train_048500", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "bad5afdf3f0741233680c5af03d76bf90614326b"} {"id": "train_048501", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "21874300007dfce615ebe8a31e18ea1521173a64"} {"id": "train_048502", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "8db1c37c9e0e095829fd41340674ec7f515a381b"} {"id": "train_048503", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["security", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "392221535a0ee3c8acb2f87ab13e2e8fcae5d229"} {"id": "train_048504", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "6d0e306ea444320f399f7a609bc64b6f698936e1"} {"id": "train_048505", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["performance", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "779b38eada577d624574e63b566676219ea0e932"} {"id": "train_048506", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["governance", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "6405ed3b4761c5f56c93306020cd1439e3c587c8"} {"id": "train_048507", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["data", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "b3010c7e79661f1442f378ce42fb51253845edb6"} {"id": "train_048508", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "024caea44d27b2c48f906b0bb8a1aecdcde459c9"} {"id": "train_048509", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Python", "moe_experts": ["governance", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "7f92eccf8f911441e21960955dac028e32d36610"} {"id": "train_048510", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "6cdd0f82ffa11fa9cbc68ad13bea59505edb5f1b"} {"id": "train_048511", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "91256f7969bb765b60284e8b08710de18f114c7a"} {"id": "train_048512", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["governance", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "5fb04dfbfc20a1ec48f702f9505021d85256f092"} {"id": "train_048513", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "ec520fc7d8bc64f61c55ebb3ff922e4be2bbac32"} {"id": "train_048514", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "6bd1574954480eb640b80015aa15ac07b44af532"} {"id": "train_048515", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["governance", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "7bec5ee83a0e7963c363026cf84b1dc2c9b2d6a0"} {"id": "train_048516", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "1a247e43a1e3d51ce34bfcd9c9cff89ba21c987b"} {"id": "train_048517", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["evaluation", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "a29a68d2bb5c4f20f60b93305f9d262f3c6a03d4"} {"id": "train_048518", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["performance", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "8550fc3199d752d6ab29849bdac9d85dbe250f1f"} {"id": "train_048519", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "468f34fea52194dfa73c8255f557b8bec22135e0"} {"id": "train_048520", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "f42f0c350677cbedba664d6d8cb31d65b43f4eb9"} {"id": "train_048521", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "0ce2d3c62e32f67774c7c812c8190562655c575f"} {"id": "train_048522", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["security", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "9fcc23cc3a357a3ef002c91f40b263e15b973c45"} {"id": "train_048523", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "20b95cad86311b865d68b9fb39b0dd9c985a05f7"} {"id": "train_048524", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["coding", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "64264f4d3dddb49a4ab71d62305a831325df3031"} {"id": "train_048525", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "81d2346d793e7c83f53442c221399083159eea17"} {"id": "train_048526", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["data", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "b7541bdd161bd7e07ef596f21da17e657862aa11"} {"id": "train_048527", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["security", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "728ffb60f5f5697d1f0ef3b98b5f21762cb4b014"} {"id": "train_048528", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "828950b478927cb1238c7a0448fe7570fae1332d"} {"id": "train_048529", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["security", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "916890bcb5e40f6ea5d402940ffb9f7fffe13b68"} {"id": "train_048530", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "5248d7bd20bedc7f3e284c559b860ffd848792c4"} {"id": "train_048531", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["governance", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "59251769ecac3e675a043a2db1f054bfa9257437"} {"id": "train_048532", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["agent", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "5d1660fe7cab6bc6d4fd8ab54b0a7ce951d4ebe8"} {"id": "train_048533", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "0d8adcc24b9a37346b340e2a3c176ad6253d97b4"} {"id": "train_048534", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Java", "moe_experts": ["security", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "914deb730b1ff5a467cc12bdb81612c5e3f8ceda"} {"id": "train_048535", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 9,\n \"confidence\": 0.9,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "8849f292a0951269a193ff64c2fdff804f73e32c"} {"id": "train_048536", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "SQL", "moe_experts": ["security", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "0b6443f939233373810e5e12fe525269d437e397"} {"id": "train_048537", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["evaluation", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "f2d900ce1c515ddad5e4667add507dae8bbf4f23"} {"id": "train_048538", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "5b274dda0241538269c15d6732f8f949e6ac137e"} {"id": "train_048539", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "d73b93871d6cd222711a8833d1a8a4e82cea354e"} {"id": "train_048540", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["research", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "2a7dc411f37f3cabe98e65c1dded8c221c3d4f22"} {"id": "train_048541", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "1806ebe8f8da2aade4a9780c111d4f81f6becaf2"} {"id": "train_048542", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["coding", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "439dff1a2aad2180bbc67dbb7b42d63192fd9685"} {"id": "train_048543", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "e4128a92d2584c2aab50c9e1ecd2d270cbcb0c0a"} {"id": "train_048544", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["data", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "19ee3ab6ba73475970f15cfb4706ce9d144e2862"} {"id": "train_048545", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "8b705a0d5e913bf9bea8dd670af4ae9567f7a1a6"} {"id": "train_048546", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["performance", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "e31a6e98fe6002f345cdc4002d7d0cdb51d947b3"} {"id": "train_048547", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "c4176b316f68f0dc65b3696a0aa67537933d84ef"} {"id": "train_048548", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "d9762e567d6433c7fd7baa6c73e2e630ee50c838"} {"id": "train_048549", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["agent", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "09d489428d2ab12b9fba0f08f79bdac0daf863db"} {"id": "train_048550", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["research", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "e9f8b9bce1422aef984d8bae9f3a1306ce202f8b"} {"id": "train_048551", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "6bfd73ca33a7ca9cd487096b7a51fa0099a3d882"} {"id": "train_048552", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "fcfe53791b9be96720839478215a94cc3c5c3615"} {"id": "train_048553", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "e666b2d69e8047fca064f1721f3123cce93931af"} {"id": "train_048554", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "e21505d844422d943b149b7d7f40ea54a715269b"} {"id": "train_048555", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "4cd76a0f88f9fff26ceb92fe563eabefa29fa109"} {"id": "train_048556", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "6ab66490760f6f9d3b88edd5b8b25d55b38f4d5d"} {"id": "train_048557", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["coding", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "1e76f42cdb5fcc5eb9b2a8dcc890a9e0df36882f"} {"id": "train_048558", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["agent", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "7e4efcb5ef97a098e702a1f21ce5dfffa06afc6f"} {"id": "train_048559", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "bcac6d0fbee6cf2a0b4fa1b7579788ccd4784bbd"} {"id": "train_048560", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "d27e9048da693de71386431273afe93be0e7a366"} {"id": "train_048561", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "f49414a44f032bbde064db94211ba3795be7a659"} {"id": "train_048562", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "1878141ab7426c86fc4a629039a33c87119e02a0"} {"id": "train_048563", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "a92a1ce538a3b1090939668a545fb2f8ce8e85e1"} {"id": "train_048564", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["agent", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "8c7c4c1539d9000f90080df444866760cf0f2b2e"} {"id": "train_048565", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "C#", "moe_experts": ["security", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "450650263df7485fc8bbe8d5f0c9c9c4f9e7303a"} {"id": "train_048566", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "6622d0cda615719625bc98e33aa34fe794cf264a"} {"id": "train_048567", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "f1be5839093cb34acd646b901b3306dfbb06e138"} {"id": "train_048568", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["data", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "f0a72273b93aabd026fd101741060f2f1ca1fb06"} {"id": "train_048569", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "97e3e673b7f15fa2a2a186468c94e0cabc00997f"} {"id": "train_048570", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["data", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "03dff8809e3af74f7b588652fb708470d07b5485"} {"id": "train_048571", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["security", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "ea78b056ac3c03463d82ed7cc60af3851ee63fe7"} {"id": "train_048572", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["performance", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "15f6de3db70f642b32e6ab1d5de6cb1a29f49f97"} {"id": "train_048573", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "7b5043b67e49f8d3031b8ad12ef730e9f4c0a142"} {"id": "train_048574", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "30941d6881b540e71768f4e567edb800ea1a64c4"} {"id": "train_048575", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["performance", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "2cefe05a35d9bb2059040bfef551eb33a5dfb19a"} {"id": "train_048576", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "11522dda8aae6e31bdb3ccfbde0797a48e1090a0"} {"id": "train_048577", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["security", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "0ef3447b7ffe75216fbf11240a641fc7b8dd7046"} {"id": "train_048578", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["coding", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "ce2126bfd7d97c5b4ce02326a104396ce86a9936"} {"id": "train_048579", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "e8a421c107679864f613d238614f4f9b5d907266"} {"id": "train_048580", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "07ae355996359608391d43054ed1de743f079f95"} {"id": "train_048581", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Rust", "moe_experts": ["performance", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "e4563e512d0060dd1a8c4eadea57c81983ac7dda"} {"id": "train_048582", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["evaluation", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "ff6bf48bb8c6a63444d0005e1823ccc599aff13e"} {"id": "train_048583", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["coding", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "f7c566324f62f0cd6d1dd46e7d225b9b70b152a9"} {"id": "train_048584", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["research", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "2e10456a6eedbc957f819e5208e2181647954817"} {"id": "train_048585", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["evaluation", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "923febdca577b74f2e0a0b2a13e98da1b2c85787"} {"id": "train_048586", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["research", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "447755180261688e39a205f46dbd789b81d54e13"} {"id": "train_048587", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "8e1017c5215d5c319cd4d07281a2eb63dbbf3904"} {"id": "train_048588", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Java", "moe_experts": ["governance", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "3a8881cd02274169dff9dd4f75ab246991995b2f"} {"id": "train_048589", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["research", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "d8b2e05e5d3b240e4867dd90e44bb6afee4d49fe"} {"id": "train_048590", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "542a94e2cd9fb09558bbbee1417d28ceb37ac48e"} {"id": "train_048591", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "SQL", "moe_experts": ["security", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "f8a1372006749b305a25ec64e4ee6d7b99ba39ab"} {"id": "train_048592", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["security", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "cc6901d324f723e4fe22d0b19c451aad0d8dda2b"} {"id": "train_048593", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["performance", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "776631ce1d3e354bfbfe814ec41a33c27e3057e7"} {"id": "train_048594", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["agent", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "d47514ebec3fbc5eb01b247bf3744b3e4244e621"} {"id": "train_048595", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "dea6f93a71c3ab79357fab396e4629b94ab2df38"} {"id": "train_048596", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "7092f91d9ce01453f0b16c311e6f36093442ed7b"} {"id": "train_048597", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["security", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "b99ab6b66c3cd1c43905e181b6682a0185355b55"} {"id": "train_048598", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["governance", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "c5ad52f96d974652e7f33a63e3bb6b167464d3b4"} {"id": "train_048599", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "86cb519742cfefa34bc17dc6a0048318e658546f"} {"id": "train_048600", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["agent", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "4ad529451c19cfa0b5afa5b4498b864b57585d5a"} {"id": "train_048601", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["performance", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "3b5d05c6f522d7bce1901ab151f62ca6f9f39fd3"} {"id": "train_048602", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Go", "moe_experts": ["agent", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "69e195f6e882b911d37eba0f7fb5fb559c2da9b9"} {"id": "train_048603", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "33df3fde93e3e776e9adaf31d77e0546442971a6"} {"id": "train_048604", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["security", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "aad9d5b5ec22c346aef8536996d2eaa4e8b823ac"} {"id": "train_048605", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["performance", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "427e7c41854dddc4077c2f3770d1fa2a2707cf5b"} {"id": "train_048606", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "bd12f9b46489ac8edfa806efc17bac0d1e1e601b"} {"id": "train_048607", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 9,\n \"confidence\": 0.9,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["performance", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "dab3b3f5c945c52733a046ec2c36c26c15ac0aaf"} {"id": "train_048608", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["security", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "4508573d501354eafb9f3bdcd97af5561a2934b6"} {"id": "train_048609", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "545eba322603c9e5cf50876574a46453c6f1d68d"} {"id": "train_048610", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "d01e53a6ee54ef64619b9e66cba9990120887fe8"} {"id": "train_048611", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["data", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "d61010e6c71899dca870be814e7900e3b666a780"} {"id": "train_048612", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["research", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "9cf6efd7fa1264172472c586f253d763ebdd6487"} {"id": "train_048613", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "a2f69b71141da58b0744f85f0873bf419c38adfe"} {"id": "train_048614", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["coding", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "0edf81129f25717d16a5ad811f57104899ff2de9"} {"id": "train_048615", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Go", "moe_experts": ["data", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "97bf45cdcfc0ea9a34991382bed43af3a191ecd5"} {"id": "train_048616", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["data", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "4fe8cdb5768aedb9f358075084941c2c0c0842bc"} {"id": "train_048617", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["governance", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "1ce215d6fddf32f07f107f38ec048154babef05a"} {"id": "train_048618", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["data", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "7e722a5492e54bab000f0951ec8a527fd8fe14a5"} {"id": "train_048619", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "38e76779466257e1c7b0dd04e5c77cba47700ab5"} {"id": "train_048620", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "e3e62c6e31344b54cb0405a8798d4e9b6dc5dce1"} {"id": "train_048621", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "2165f5d9e093d4f72a661d283b713a3591525ef1"} {"id": "train_048622", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Rust", "moe_experts": ["coding", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "2571590d066a2fd51196db0225981a9befe64fee"} {"id": "train_048623", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n cost = 0\n for _ in range(budget):\n p = plan(issue)\n patch = edit(issue, p)\n ok, _ = test(patch)\n cost += 1\n if ok:\n return patch, cost\n raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["governance", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "35aa639c03ed1633860bafaca052a0b9b984e663"} {"id": "train_048624", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "d9b9fca7e038cf8b1faa1fc15c3f6319500469ae"} {"id": "train_048625", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["agent", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "0f7e72a07ff727af8b145aa3d05f9adea569e5d6"} {"id": "train_048626", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "f92875c875715ab6d2c137a14c180800200eab19"} {"id": "train_048627", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "534910804a21a6ce7af38c9abc84fd3d7b21d628"} {"id": "train_048628", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "0302afcbca45d4961b34a940726c0e8f52ae1472"} {"id": "train_048629", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "e0e5f1da45e2edf76c81e1909af86e772c8a6540"} {"id": "train_048630", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "1cef0751507a72e1114b4a8bed2a5731e9a3e8f6"} {"id": "train_048631", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "b4ed86569643508e04a0c4445160cff29c48079d"} {"id": "train_048632", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "20855f66e66be4a741df452146a7507a11473513"} {"id": "train_048633", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["governance", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "c1ff560945c4186103cfa630e2bd5e728b117138"} {"id": "train_048634", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "dd504db6f25e7ccde3cda45339006f0be05eaaa8"} {"id": "train_048635", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["research", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "6d840882cb05ddcd28d28bbd076ec96fa5d134c4"} {"id": "train_048636", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["evaluation", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "7df441dec679af77bb33a2b843157e3f2e72ba99"} {"id": "train_048637", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["research", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "02172eb5e6ee2cc5ae142eff462683a5b4767eb7"} {"id": "train_048638", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "f49414a44f032bbde064db94211ba3795be7a659"} {"id": "train_048639", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["research", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "2260a9a8e685c1e2ad7c3a3621ca18f831e5b199"} {"id": "train_048640", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "c7f143e83515ef519d4504d52c852dc633c1991d"} {"id": "train_048641", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "acc7c0ebb03ed8e366eb1c416efc8dae65f72318"} {"id": "train_048642", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["research", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "5516ed0ea400204216756f398212ca835c56f0db"} {"id": "train_048643", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "392686273cd4400e7f504bbe3ec48063615c23ef"} {"id": "train_048644", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "4efc958436354093bb6fb6f3e970960ebc9f5f11"} {"id": "train_048645", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "b9228577d212b9f16e299f490e6e889b60dafbb8"} {"id": "train_048646", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "2a3e134edc5c1c61e7b09cbcbb864a2b4f1b23e2"} {"id": "train_048647", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "13a3e81d2dbecf5478e2583bd4b4081f6120e158"} {"id": "train_048648", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["agent", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "79bd784ed9cb863781889d7b56263f7de1607fb1"} {"id": "train_048649", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["data", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "4f47e3cec9c9ea7735240f06ca85fc5227723818"} {"id": "train_048650", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "SQL", "moe_experts": ["data", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "65d666c730e3920275dacb5eed4aee35428ff585"} {"id": "train_048651", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "C#", "moe_experts": ["agent", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "9b3b892214eeb38c1f48872054146a5c3e2266e7"} {"id": "train_048652", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "8a10d9487e962fc5d16681942297f809123f495a"} {"id": "train_048653", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["data", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "fafd3c206f7f17212a79a6fa04ce4bd377aa77e3"} {"id": "train_048654", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["data", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "b7f28158106a9db8a7c2b221ec1813f5ed9cde64"} {"id": "train_048655", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "SQL", "moe_experts": ["security", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "f990c3e314da635b2ba55a8fe67188ba3db194e8"} {"id": "train_048656", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "9adfbd4590dc0367c7a37ffbd4bd696007cf0a7b"} {"id": "train_048657", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "a5a483bda7cf4d5191cd5d917d8361735e88cfcb"} {"id": "train_048658", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "C#", "moe_experts": ["coding", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "bfa670b9c2978994cb79a13a6d005cd0047c3355"} {"id": "train_048659", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "ce42a10765ed478c019099418a48be8d5f9dc2e9"} {"id": "train_048660", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "3a8f99da91fae971c69365924400e2feb7cdc408"} {"id": "train_048661", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "66e784ea369b5a46a0562d705f411bb27f64c78f"} {"id": "train_048662", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "dc45df23b49c682bd8d8b03bf3e37e08ff534653"} {"id": "train_048663", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["performance", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "71b81014f75723464196bdb55f40c3927f7fc2a2"} {"id": "train_048664", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "a1283754789ddd5ebb8c31c1323f557515de2af0"} {"id": "train_048665", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["governance", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "1bec3c9145d459f01827511a63145fa91ccd645a"} {"id": "train_048666", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["research", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "c6f0a02b94d6e947b80083ae8d34a69623650476"} {"id": "train_048667", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "C#", "moe_experts": ["governance", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "cfc8897db50875dc1522a706a41927d970d86fd2"} {"id": "train_048668", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["agent", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "d5580800a54e01086405fbd5728665941105484f"} {"id": "train_048669", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Go", "moe_experts": ["coding", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "a18a693d0e978fec6c51e0ac947a8d5738a2160c"} {"id": "train_048670", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Go", "moe_experts": ["performance", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "4352c6e38b891cafc67bacc251fc07f190934b8d"} {"id": "train_048671", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "2dbae08f7734db66131ea4fe18f5aca1a82bc08f"} {"id": "train_048672", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n cost = 0\n for _ in range(budget):\n p = plan(issue)\n patch = edit(issue, p)\n ok, _ = test(patch)\n cost += 1\n if ok:\n return patch, cost\n raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["evaluation", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "72bf017787bd488ee367910a53914fcde812b5d2"} {"id": "train_048673", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "354083d148cc9dd93976843f62ef5d65f586acd2"} {"id": "train_048674", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["coding", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "121ddfd9b0f3a81e432f834a8c852a2333a6cb93"} {"id": "train_048675", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "bcae1d93e5451e896afe72547876f8f006f186eb"} {"id": "train_048676", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["coding", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "8320f99346cf0ad13c14ea538193f1ee19cd7dab"} {"id": "train_048677", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "a3cf646c768204ec3c5aded85b4707d1cb585724"} {"id": "train_048678", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["coding", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "dbd2804c059d2b55c3f582d7e100da957793b54e"} {"id": "train_048679", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["research", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "5908ec8c26b52c257898650cf5402e7154a34c05"} {"id": "train_048680", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["research", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "80e7d050c0fc5c02bb6ded3f66513c4f39db825d"} {"id": "train_048681", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "eab556d42943bed383a13b09781870200bfd5301"} {"id": "train_048682", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "9b75ec8c153182951dfd21c277b9539dbd76d210"} {"id": "train_048683", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["data", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "47a04e503eadd46aed40d79b6882be900ab1e270"} {"id": "train_048684", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "fc264298ab5466508c8f2a648474b06df4f12b9a"} {"id": "train_048685", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["agent", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "db155495dce08f127ccb3b31c36f36a06809b821"} {"id": "train_048686", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["evaluation", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "f4ee5b4debc8bd5d6ba94ffd7c5c6366968fd5af"} {"id": "train_048687", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["security", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "11d52b4734631b0fde38231f4c06156a0f19df01"} {"id": "train_048688", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["coding", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "130dd659e277443f67f67a23656db897a93b7834"} {"id": "train_048689", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["coding", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "08d1ea526fbb5e6fdfbd1d02faf5ba0df45196c2"} {"id": "train_048690", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "409823b28396a302994eb18626fd68f105c08204"} {"id": "train_048691", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "b556cadf8bd1454d7f9e59ecfd3ed809da07badb"} {"id": "train_048692", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "87138d8486a55a82a0b209709ff17ba47fd1e66b"} {"id": "train_048693", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Python", "moe_experts": ["evaluation", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "399737f25d00b0069931f06baebc0f742a9ae033"} {"id": "train_048694", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["coding", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "2514ed168a5e16cc0ab325ff2dbe9fc0ce2ff028"} {"id": "train_048695", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["coding", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "9e76a4be50eadfe4a97a8e3f4d1ba38050a2c8ce"} {"id": "train_048696", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "5914ed858c351a252d9e36606ce612ff33ce4599"} {"id": "train_048697", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "cbc2354740168569b8dbb154035bed52ae35e3d8"} {"id": "train_048698", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["governance", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "3f2daaa9da64bb474faf682efd64efbe3447fa88"} {"id": "train_048699", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "3ac01de5913705a54369fa4d192383e2b2f2c9d5"} {"id": "train_048700", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "8d5523318ca3b8f4e6a3d04bd627045720185fb8"} {"id": "train_048701", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "SQL", "moe_experts": ["governance", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "89d5bb1bc0debd238487c31d2f6362e7846c35c5"} {"id": "train_048702", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "0d6f9765b53ab90cf281a231abc6122723585834"} {"id": "train_048703", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["agent", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "adfd68f2cc2db804b182ba1ddd9c0e033d2d53bc"} {"id": "train_048704", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["agent", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "728bcafb345381d32094e4fcc901490250804379"} {"id": "train_048705", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "caa1155ba3ce5b5b7bce7c66cbafda6cb6d657ec"} {"id": "train_048706", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Python", "moe_experts": ["data", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "a8e968982663fcbf14699b954df55103d52c3f69"} {"id": "train_048707", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["coding", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "f0914ea987c5aca84aecd2f5e195ea29befa6909"} {"id": "train_048708", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["agent", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "a7c564d2a0b091c9e135ec8e7663ddc827a8a31f"} {"id": "train_048709", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["coding", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "82bf81096ee2744bc08b1dfbd802eb58be8e3de3"} {"id": "train_048710", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "3b2c6df283812962397c7788d8f96650ed952c54"} {"id": "train_048711", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "23b7a34725d1addbf6510d0460a6d9f4fc500391"} {"id": "train_048712", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["performance", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "d17e04da91d44b998ee96414f657aae01b596549"} {"id": "train_048713", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["data", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "e0eb024413b6779bc73e3f37a03aa74298fe908c"} {"id": "train_048714", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "3e3cf5976642a299d18c390e65c21cfd178161a8"} {"id": "train_048715", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["evaluation", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "2f6e578b42b5c6f29c15ac5cf692fa5549dac08b"} {"id": "train_048716", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["agent", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "be68a1e3f4c7c35d319b0f7d96c8101b07d07d26"} {"id": "train_048717", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "3f0fd682929496792bb7768058d9bb95a53b61ef"} {"id": "train_048718", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "SQL", "moe_experts": ["agent", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "a05a3280fb14ed280de5e5cca67d6f95dea4d65c"} {"id": "train_048719", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "d1a84ab46ba3ef58edc52a611b409dfd8dbc3464"} {"id": "train_048720", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["security", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "786b80286c309cb0c15b00d510a249ec7465c1c2"} {"id": "train_048721", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["security", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "22ce7ef0ead00525f591d26081b2bab1fe470da7"} {"id": "train_048722", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["security", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "fd34a7d72bf686e8e46609118fabd5fb1cc9f396"} {"id": "train_048723", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["research", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "2ea2075bcfa1eb25860325f43761feb5bb135766"} {"id": "train_048724", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["data", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "21ba9a0ac717e06a7fceef3ab1a202d926ec137c"} {"id": "train_048725", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["performance", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "9b68b94d09b4c478a2c12f13af4143cecd968e28"} {"id": "train_048726", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "eb2f3869a90df15c660db24e4f6cee8f3193394d"} {"id": "train_048727", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Rust", "moe_experts": ["agent", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "243a5fab2c1f8f51b82494d7b03c10cb82ed373e"} {"id": "train_048728", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "d877af7fdc864c145ca1dde4f71f98d77ccf7a39"} {"id": "train_048729", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["data", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "0ef015aae168b3f9f5f6532bc408d8d4b32f869a"} {"id": "train_048730", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "SQL", "moe_experts": ["governance", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "16e2614d53ac4707a4ce46a92ebefc665e239f1b"} {"id": "train_048731", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["security", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "69c624b1e45658cb04eb8801d613386d72be7683"} {"id": "train_048732", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["data", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "7081f9b5811b79cd1ba6853ad375b54d3f43f0c4"} {"id": "train_048733", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "ad819396e0665c3e1c1be1b6e451cd741cccaa67"} {"id": "train_048734", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["coding", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "c9a0f481cef5921900e7a85159d863f5666ff5a1"} {"id": "train_048735", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["security", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "a81dcf7cd6bd860999464579ddbe8c15f7033c57"} {"id": "train_048736", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 9,\n \"confidence\": 0.9,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["research", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "b4e15662da28b12a5e702b220f6f5160db4b3073"} {"id": "train_048737", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["performance", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "3cc058dc50271eb2fa8786ebbe9ce20f0118883e"} {"id": "train_048738", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["evaluation", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "b9948c99e40bcbdb8b3eea8a51858ec52b7f923f"} {"id": "train_048739", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "2303831f7c8f2bb6aeb8a4eb8b325c8396c9b416"} {"id": "train_048740", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "745b87d1d2c8dac9da5873e40668ae4b78cfa51d"} {"id": "train_048741", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "8e3cecf94e3110d7a4e437f2264bcc20f86f52c2"} {"id": "train_048742", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "da2aa776c5e3ff08a98e07be80d34030c8073d4f"} {"id": "train_048743", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["governance", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "8b17714d4a224741faa5e3245172e4af3fc6ae1f"} {"id": "train_048744", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "0f491a546efb18ada634fa6afddf4b0a3e030ebf"} {"id": "train_048745", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["coding", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "08eef8dedeb2e60b7bbea5005af36693e5f55a38"} {"id": "train_048746", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["research", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "1850c90da7c75f5d0958bfecadde00e7474e52e0"} {"id": "train_048747", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "4e7c54fee0966595aa9839cf149035c1d0ea255c"} {"id": "train_048748", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["performance", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "0d71baa7a3601251ae33487329fcc89cc3dd16d5"} {"id": "train_048749", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "0f99de0fb16a3197cbebf0b287b8efda19adca98"} {"id": "train_048750", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["coding", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "8d9d2d24abdf50f827584a0f2134d5ee6c63ce9f"} {"id": "train_048751", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "847f2593d29cb91d9d3be5afaaa9f04171a466b2"} {"id": "train_048752", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "c98d0ebcef026083f66c93af9c017c27be40fc9e"} {"id": "train_048753", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "034f653caf7f7f8fb42bbdba60c4537cca657122"} {"id": "train_048754", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "7f5f1b83f9632fe5e93f8432b869251de73cc2ab"} {"id": "train_048755", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "c49b5f938762bade659b12969d4f058cccd8fe59"} {"id": "train_048756", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Python", "moe_experts": ["governance", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "e9cb4d4cd8de6fcc4d5cc2ab5c0eab42e7ae7c1c"} {"id": "train_048757", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "bd25d1bb807db30e2fde4c8c2e33a98906310155"} {"id": "train_048758", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Rust", "moe_experts": ["coding", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "7c90998d0030a3e0e4ba65f6c0706a342a8b45e6"} {"id": "train_048759", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "3685e0ca8d7f9785bee43d638f6c4dcd1b6b418b"} {"id": "train_048760", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["data", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "a1e91fb57456720621b799cc80b6c4ef58784f47"} {"id": "train_048761", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Python", "moe_experts": ["agent", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "23282445f554748e9a43d53b8fb5b5ea64c33807"} {"id": "train_048762", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "ec869f64b3f6569c25c92937b53f85e6087f444a"} {"id": "train_048763", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "C#", "moe_experts": ["research", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "0626431dc51c1871dcf6333af2029bd5de1cfafc"} {"id": "train_048764", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n cost = 0\n for _ in range(budget):\n p = plan(issue)\n patch = edit(issue, p)\n ok, _ = test(patch)\n cost += 1\n if ok:\n return patch, cost\n raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["governance", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "a2b377e4dc6e3df99472d8878a7f497b0998d590"} {"id": "train_048765", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "ef28d9178d8a1aab514824bfa7fed1f76e3c49a8"} {"id": "train_048766", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "5062806ef2ec17cf7147d902f87dc5eee18bebcf"} {"id": "train_048767", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["data", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "f082a1382330353054ab951e8d46006ba1d8a763"} {"id": "train_048768", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["data", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "4982a5c546747167171e6cac20a0485a6acfad00"} {"id": "train_048769", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["agent", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "8019f022d05700e1a86fdc57757840802bae7d98"} {"id": "train_048770", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "83272df83b18e8a9b04083e07e18594657793e37"} {"id": "train_048771", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 9,\n \"confidence\": 0.9,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["security", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "0f6ad8b6a539bd466f089065b46bcf61fae7d1be"} {"id": "train_048772", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["agent", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "e44b4c8872e6796fb0c3b2e7b4116f7a1df0a4ab"} {"id": "train_048773", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["research", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "d8766015e370cfcec7c6c578eb5e227b9a99a7ac"} {"id": "train_048774", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "d6b2d5cdcfd44c0adcded8916c59389254b4f409"} {"id": "train_048775", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["governance", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "f87232cd658c0a3cfd8af758248fa96cd5e5cfce"} {"id": "train_048776", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["agent", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "a82757dc11b6e2f012527d938e091e40bbe8aab1"} {"id": "train_048777", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "8842ce66d36ad89bf9fe8684430e26b9fe771a17"} {"id": "train_048778", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "3f1c5e5695b3aeb50ee795fd124960e6272e7946"} {"id": "train_048779", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "12d72501c3b2948ac2d225a2f704fcd639186f7d"} {"id": "train_048780", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["research", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "251fd8e35b8df3a96c93b619f58c8dfcfaa32781"} {"id": "train_048781", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["security", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "181e91444ebdb37109aa85cff4df50df608e89d4"} {"id": "train_048782", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "608020e68c4651ffec513f51fdeead8a59464642"} {"id": "train_048783", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["performance", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "5233288905d1ee477c217b61273c3f85f00ba1d6"} {"id": "train_048784", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "47bb36cafe948ed1e081659e220167b0e802a2f4"} {"id": "train_048785", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "db8ce1e3c7830221fc893ea089630b19a02b263a"} {"id": "train_048786", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Bash", "moe_experts": ["performance", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "079047e1ab7bc10ee7be630eb47cdfa4e7b59b94"} {"id": "train_048787", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n cost = 0\n for _ in range(budget):\n p = plan(issue)\n patch = edit(issue, p)\n ok, _ = test(patch)\n cost += 1\n if ok:\n return patch, cost\n raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["security", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "09d65becb6458c4ca001198e0492295b0fa3b805"} {"id": "train_048788", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "2a0b714ffb8dba5f9c0ddd11aad40640a3a0b4cb"} {"id": "train_048789", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["coding", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "234058a11fd0fdb1a307bc8c0f3995c8fa775b91"} {"id": "train_048790", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["governance", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "f018518258480b0c96e94387d5850785af2ab9f4"} {"id": "train_048791", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["evaluation", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "cdca2f1b9c93b8c81022fdcc4dca4fac44f09388"} {"id": "train_048792", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "a925d59d5c78870d1ebaef4f25a4b1d4ea68d51a"} {"id": "train_048793", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["security", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "100a5cede9532ed8925a2140ad7f6dc59f08003b"} {"id": "train_048794", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["evaluation", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "fbb2805f05bc16ea0775ce5063a4089686aa551b"} {"id": "train_048795", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "210a2ed9b844bdf8e99500cf402b7aff90684513"} {"id": "train_048796", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["research", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "3aa225baa0cd7f283a530c4357404c9ef84fe2be"} {"id": "train_048797", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "e619f1a65303b4900e364b5f09182dc2c1eb9ee7"} {"id": "train_048798", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["security", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "5f8dbe09a348cb772530d2f940ccfbc2107a2c9b"} {"id": "train_048799", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["evaluation", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "a12bd8ea7d818ccf2a6cb7acf77d577ba8793608"} {"id": "train_048800", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "65d996ce87d918309595e6aeca3ce88a3750ca98"} {"id": "train_048801", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["agent", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "d6dd153f54fbd7dca4c249e65d62cc5c948b3eec"} {"id": "train_048802", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["security", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "3505aa8c8cd524072bce62d9134eafa4f9fea911"} {"id": "train_048803", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["security", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "ef1277b93b77849cb02897a44a7e60c09e8719bf"} {"id": "train_048804", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "b75e71710a8efa609f7eeca6eda1059c34b53e28"} {"id": "train_048805", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "fac0454c20e3d4e9a6cf64a55b5b8b728361ca70"} {"id": "train_048806", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "342aa9450a640d8a9d813ac0a65bfcdf33e4db1a"} {"id": "train_048807", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "20e09f2d4b0c4664043dad12951500eaeec8198c"} {"id": "train_048808", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "C#", "moe_experts": ["agent", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "95cac0748ce7f405582dd0280a5dda37f98eebbf"} {"id": "train_048809", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["agent", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "27b61053ecc757d2655a25b85a1a1d96d9c63db3"} {"id": "train_048810", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "97ddce7ece9f8e8c1515e2e8cd5c2dbe334a0d81"} {"id": "train_048811", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["research", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "c62455d25b7b30b7d15d56ff66f0b97cd5ffe2ef"} {"id": "train_048812", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 9,\n \"confidence\": 0.9,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["governance", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "d67562a446e74eebc0dbbbec93439fad66c78d87"} {"id": "train_048813", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["agent", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "a44db2c6d3dbff7ac7d16a6e6751d2d5ba4bd322"} {"id": "train_048814", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "424565d7e8db39562040afb125af8f2731eff37e"} {"id": "train_048815", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 9,\n \"confidence\": 0.9,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "4e3be9d5b0d2ea1bb7538b16173ed445e876aa6a"} {"id": "train_048816", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "b17217cedb67394c5bbdda6835e874e381df70ad"} {"id": "train_048817", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["security", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "7b45d91dec99a46a5347c05b94259c93e7420a15"} {"id": "train_048818", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["security", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "3e0b4cca56013a90214632a4efafeb79b9da2983"} {"id": "train_048819", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "f6db4b081fc99270b556920b9514c880ea52140e"} {"id": "train_048820", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["data", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "bb7f9ce36f4f7b8d23af5706cbdf1289d5f457cd"} {"id": "train_048821", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "d301d612be5e1d1d5589318fcbad81efb2184f9f"} {"id": "train_048822", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["evaluation", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "ec2418059dbc1dd82e5bda6e7c0b303844204b8d"} {"id": "train_048823", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["agent", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "d34566264bd00bc4f99d9b559fbd311c33256892"} {"id": "train_048824", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "92887e57729a397dc5f7f0acc594a20a9f917223"} {"id": "train_048825", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["security", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "c1ff560945c4186103cfa630e2bd5e728b117138"} {"id": "train_048826", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["security", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "a5a1327cbae1387934643ae308805164f1a5459e"} {"id": "train_048827", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["research", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "519b58e6318c49f315018848bfffa4184c194f4a"} {"id": "train_048828", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "8aaf01fbbf3b0c9d41069482f6ba7df16727144c"} {"id": "train_048829", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Bash", "moe_experts": ["data", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "0a48658d163330d34e0e3e15af566188dba9fa3c"} {"id": "train_048830", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["governance", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "9518bad2c82d351fd18d772012ea9969f1a5a6e4"} {"id": "train_048831", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["coding", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "aeffdc9d242e7932d60d732e9c6659921c3be807"} {"id": "train_048832", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "75150bba4e8e2461e4085d1e8ae55f0e71b7d23e"} {"id": "train_048833", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["evaluation", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "2b4475f87bc9b5e707dfdbcfbe3fc867d9e981f0"} {"id": "train_048834", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "923febdca577b74f2e0a0b2a13e98da1b2c85787"} {"id": "train_048835", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "f29428e62b652ae5cc2ee91762924b9b5b34fbc7"} {"id": "train_048836", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["performance", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "7bdeb420ea6dd790e74a38a8ba903f93e25269b6"} {"id": "train_048837", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["research", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "05a298d01cd6dba662a277e89f0c45183c825853"} {"id": "train_048838", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "f2b6170acb686c791a571573af9b373b118ea091"} {"id": "train_048839", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["research", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "23300a9890c815a3d672c6a4069bf6194d5440f7"} {"id": "train_048840", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["research", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "702a127cdc663f38aa43ef5a5d3a7b8f7d19ee3b"} {"id": "train_048841", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["governance", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "6af8e114262527759b7281d8cb8eabf48cd9851f"} {"id": "train_048842", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "2430dc4222503037e9da18a4804d7d663df9deed"} {"id": "train_048843", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "00d8494f60a36a27027eb96a304413085a03d86e"} {"id": "train_048844", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "b2d1f307a0709e33d583c2ccc3d6ed2830d74786"} {"id": "train_048845", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["data", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "5a67a93a7cbaa28b93e421c56516b9cc69845f7f"} {"id": "train_048846", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["data", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "f29428e62b652ae5cc2ee91762924b9b5b34fbc7"} {"id": "train_048847", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["governance", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "c9ea87c56df24ba6d83436c5a71623942d826c81"} {"id": "train_048848", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["research", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "308be70043dd953a0056cfcf82f2ae5977c88ebd"} {"id": "train_048849", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["data", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "deb5d50431fb09345b0eb6d9b69c1d7fc7964656"} {"id": "train_048850", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["security", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "825b735ffe52fb51700b34aea03a2486e314132b"} {"id": "train_048851", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "89d57c97a47096565dbff676d937d0d098e590cc"} {"id": "train_048852", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["agent", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "989c1bf9a2f457ec48b00761e799d35e5e536c73"} {"id": "train_048853", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["agent", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "3f72b5622ebda0cdea339c5b48c50091c7194e1c"} {"id": "train_048854", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "483c8d681f946885d4d043e04420d3f058d67c0b"} {"id": "train_048855", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "67e484ec47b691347d0519615f83306daf855ba3"} {"id": "train_048856", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["security", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "e72f29b2937b7eb2797676596f4000ef256e3b89"} {"id": "train_048857", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "c3d46be65f58c40b707999ff4b103d0d50536b87"} {"id": "train_048858", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["data", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "9c12f9b0d8d00dd24913f09023e636a6e768d3e4"} {"id": "train_048859", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n cost = 0\n for _ in range(budget):\n p = plan(issue)\n patch = edit(issue, p)\n ok, _ = test(patch)\n cost += 1\n if ok:\n return patch, cost\n raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["performance", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "a7f56c00a082d8d3ee6e9f689962284a0e8e34c5"} {"id": "train_048860", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "b608cc5c12c24c0ed5af7cb73a0780a237b9c910"} {"id": "train_048861", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "4ad529451c19cfa0b5afa5b4498b864b57585d5a"} {"id": "train_048862", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "1d66e0b2c7eaf4c163b03746362a27d92ea1b047"} {"id": "train_048863", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["performance", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "a3d8eea8034f2e5d791c7b774576faa609002562"} {"id": "train_048864", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["data", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "340d2fae05f9d256bc9197ab0506e7f67b444f40"} {"id": "train_048865", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["coding", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "6ba26cb52ee1679c008989ea68fbc925735c5243"} {"id": "train_048866", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "ae2c42f0ce7dfabb1412177c59324c60713ff580"} {"id": "train_048867", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "ed4a63031a93f94a3707afac7bf4d3a4503db683"} {"id": "train_048868", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["coding", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "5c585d2bf53bb4072b0da563f3bb23d57022d291"} {"id": "train_048869", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["data", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "00212f0aede1929407bbe314a48855fd09c0a447"} {"id": "train_048870", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["agent", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "badbd52cb6670b3179120f78d4ace323117d1516"} {"id": "train_048871", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "C#", "moe_experts": ["governance", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "e79a942a5bf23e7c96f3d39bc9832850bed558c2"} {"id": "train_048872", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Python", "moe_experts": ["agent", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "bf2e1ae34901821979126e4d9d6f7cd897ec2ef5"} {"id": "train_048873", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["security", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "3859be78d3c22c1bcd47500dfa72535db41b8060"} {"id": "train_048874", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "C#", "moe_experts": ["evaluation", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "09ff15d732cd47e24458ede24deb53248f7a9312"} {"id": "train_048875", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["performance", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "915777480f2d1aac3feaec3a0b1594dbbc7e6e07"} {"id": "train_048876", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["performance", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "2208f74cd17d9316fbae771c62c43574abaa6134"} {"id": "train_048877", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "33c5b6dd3286dc7838e74bc2c04ac7bd6fbccf31"} {"id": "train_048878", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "9feb1dc7a2376004e74102b112674e3db2903e54"} {"id": "train_048879", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "SQL", "moe_experts": ["research", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "630ad727f77dad81b325298228570843025d23e9"} {"id": "train_048880", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "7476416111778e2e46e4627b0019a7264ac6c528"} {"id": "train_048881", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "49a57e53b6bf8de92d5b5364db4c9da706e27ec7"} {"id": "train_048882", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["coding", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "773852a24bc129019c1773aa47bc4199fec22f94"} {"id": "train_048883", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["coding", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "0bd5412f282004b14783bd811986fdc53aa5558c"} {"id": "train_048884", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "78071b3ab6b92be5aa7f9d8a1e7ae2b05bfd4256"} {"id": "train_048885", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["governance", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "a01653ad6ca0c219d6114c188b870c09f001df81"} {"id": "train_048886", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["research", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "b198047b54f0cd663c82625662119091be871a2e"} {"id": "train_048887", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["agent", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "3171d39d7afbf66e23fbcf4b8bc2bc2cb6dba98b"} {"id": "train_048888", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "0c76bf6e5f544a985037c2124264dc5f9922f791"} {"id": "train_048889", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["evaluation", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "88cbadf71a3b52e5b6740269099074740aa2a974"} {"id": "train_048890", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "182b6b0ea544b4ef3ed9afbecf916c3aa8b0d37f"} {"id": "train_048891", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["research", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "0c2568bf2e5544605ba41aea1c280c38febd170c"} {"id": "train_048892", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Go", "moe_experts": ["security", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "f60e42b0ed946bd0f0ecd045ee3e37546a530916"} {"id": "train_048893", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "8b1d06b13f795f77e8cd43e80d823e410d727d6b"} {"id": "train_048894", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "01a9c279a14844dc94325ec9e2c43e890441dfa3"} {"id": "train_048895", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["research", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "1c391b43a944d8fed824e252e25d23a74cac7ae0"} {"id": "train_048896", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "9d7196f91776b4a14da930f7d5073feab8eceb8d"} {"id": "train_048897", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["evaluation", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "060171697243859dde108fb228be31fb62c889fa"} {"id": "train_048898", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "ced4b2456cbf5d8c68bb2f0e2124246b549c3db3"} {"id": "train_048899", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["data", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "4f1141212e555e632fe924eedf289984a8556b53"} {"id": "train_048900", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "0db2cc150471572c94c3b81c20d57d7926c3e73f"} {"id": "train_048901", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["governance", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "d6a23a74cad7604001a8f3e03f0f57b512205d62"} {"id": "train_048902", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "c2996ff23231a0abcb30598f0fd52d2a9251f731"} {"id": "train_048903", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["data", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "3c29c3a0641169ee0d7d3b65e106a24f1f974aa6"} {"id": "train_048904", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "aa8107936b2f93c4e305c95bf96e7afefce10882"} {"id": "train_048905", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["governance", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "5132cd1ec863253010a6b009d36e2a26db8029eb"} {"id": "train_048906", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "00e9c6f144ea8d04b03ffd57f8ba86d5abe4a813"} {"id": "train_048907", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["research", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "c8f5385d27ab45ac378e30154585752b62d5f336"} {"id": "train_048908", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["research", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "0b7bbd1869c9c76772a365dde62224ebc3030e16"} {"id": "train_048909", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["coding", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "bddfb5559431d0e9e2c0eec66804dd06d156a155"} {"id": "train_048910", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["security", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "94c0560ec0e97afc334c3bb0801db4fe9b646036"} {"id": "train_048911", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "f28f611e1a4bdfc9951d74914138d1c6ebc90f34"} {"id": "train_048912", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "e2f2bbbe4feb4056f587bf5506c05c723135692b"} {"id": "train_048913", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["coding", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "ab856ad8681b735f4c1a55e8aef5cbd5a47afcf7"} {"id": "train_048914", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "ad4849b876e119ebbfaf8dffcab9b6bac394090c"} {"id": "train_048915", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "5180116ef1bb6e73744eb91c14318f3750ea2ec6"} {"id": "train_048916", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["governance", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "cf263be34b49e68bc6312d884bde6960eaede738"} {"id": "train_048917", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["research", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "1b82ef66f449e84e40b044d33aa6ab895d984e75"} {"id": "train_048918", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "c9d1de993f616623169b9dd14f8ffe9c4f6b8497"} {"id": "train_048919", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["performance", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "5466df7feb7e426f2cb6752da1bb7b208eacd174"} {"id": "train_048920", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["security", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "94cb3d728d9dabec10bb8326b92802dbc24f7860"} {"id": "train_048921", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["data", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "5d126915942406d571a019f9f49979ccc4e40511"} {"id": "train_048922", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "cec172dfa3dc12f0136923e5bac01b15030830af"} {"id": "train_048923", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["data", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "8f7a6aae00837b2913445c41dfa4dc7df1619217"} {"id": "train_048924", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Java", "moe_experts": ["agent", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "7822553c4e576513d4f5940cb50500f42e84ad6d"} {"id": "train_048925", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["governance", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "9549f5a6735cc8fcdf83d7ef9f51bc693e653f59"} {"id": "train_048926", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["research", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "0342d639220270e8d2f4c77a6b8b705c1529d1e1"} {"id": "train_048927", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "2ed5935effd4999fa039539a5472c802977163da"} {"id": "train_048928", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "ff0d96bb8b3fead610cb00b6cad2b0e28c95396a"} {"id": "train_048929", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["data", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "bac68f25992f49344336b483433c8a69cf3d371c"} {"id": "train_048930", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "acc090d1986295945fbf280d86f0cc842f82b4c8"} {"id": "train_048931", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["data", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "2ed4ca1dcf9b4b7b07c908505b47d46d5008a6fa"} {"id": "train_048932", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "dbd3f8162f4ac5310355053e437c3eb5b5a2b82e"} {"id": "train_048933", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "e5deef33612a16cafa782c6508884a8549b0a497"} {"id": "train_048934", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["coding", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "916890bcb5e40f6ea5d402940ffb9f7fffe13b68"} {"id": "train_048935", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["governance", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "48acb8a5834059ce2ac390015b01d5cf65b6d47c"} {"id": "train_048936", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "46115203369fb54e6fd64d11bb7d11cf18531d55"} {"id": "train_048937", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "c1f42b56675e7a3fcf2de25924652fd6fe0ebd6f"} {"id": "train_048938", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 9,\n \"confidence\": 0.9,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "bb26df5f44f6adab9fe1a6c9f433e9af30f0590f"} {"id": "train_048939", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "3c418f65609641d9b2811eb17ae789c4f5d04728"} {"id": "train_048940", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["evaluation", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "daafa227f93f8888c4ddcda4927706e42a56ab73"} {"id": "train_048941", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "0a0d1b32053a63163d38c55ef45483be35c4173c"} {"id": "train_048942", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "6706d52c4eced4c3b59525d425ccf958c36efb78"} {"id": "train_048943", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["security", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "278923d979910313d5087c5471c392c3bafe7a00"} {"id": "train_048944", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["evaluation", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "fdb4f9313f77e51e10535b3b2cf567c01dcd40f8"} {"id": "train_048945", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["agent", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "643dd91f09f8b50c9b9915f411cf3eb2928fde29"} {"id": "train_048946", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "1a52062bec86a73c8c07b8e4581cc5f49de0b1d4"} {"id": "train_048947", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n cost = 0\n for _ in range(budget):\n p = plan(issue)\n patch = edit(issue, p)\n ok, _ = test(patch)\n cost += 1\n if ok:\n return patch, cost\n raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["security", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "d52d73df8076eea629e68901cb8bc87c4fc99b21"} {"id": "train_048948", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["governance", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "98186d2aaecaa5930403edf020453e55d74f06e5"} {"id": "train_048949", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "14003e8d5e9eb98ebe68885a4ffc1b299df9d0fa"} {"id": "train_048950", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["performance", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "cb113059e5df6633bd3029ab5d23fb333a1373cd"} {"id": "train_048951", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "19bff1069dd84304089908959a1c2195de1b25e6"} {"id": "train_048952", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["research", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "8f7a3628a73ea1b612eedac6265b80a87345290f"} {"id": "train_048953", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "94f8ce92c83faa01f908557971f7e8b6812ee225"} {"id": "train_048954", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Python", "moe_experts": ["research", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "9efd1a454f7404099acfec4dcd088a8b4bd0b328"} {"id": "train_048955", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["agent", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "b719137f3c7a78d7149e092a08d9de43e2c5fd09"} {"id": "train_048956", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "81f3a20ddaedcce4fabd1430059844be75b8e102"} {"id": "train_048957", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["performance", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "298fc592bbb50803f4f8c905ae95315fd573ad9b"} {"id": "train_048958", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "SQL", "moe_experts": ["security", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "91125ae2022236376df9a92a10c927b79b6379be"} {"id": "train_048959", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["research", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "5c5461c46444e9c063c1d2c8f6586560ce05ec57"} {"id": "train_048960", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["data", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "962ed4a60f3d11ed6778babcbcb5582f3df75bd8"} {"id": "train_048961", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "cb3eb2679e6b9565e2924dbfe1fd844c435f0b90"} {"id": "train_048962", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "e3e61ec2c22d4885f70166cf3a6b18383d1a92b8"} {"id": "train_048963", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["evaluation", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "1329a56bbc2f0e52af34a6c7863b4e919b37dcf3"} {"id": "train_048964", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["governance", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "1ec0a317d9cc36f192b35413cfcdefd557dca4b3"} {"id": "train_048965", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["data", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "a7d6c29d04649f519965f86d07dea746490fd77f"} {"id": "train_048966", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "44d49c373d545450c6e0ea43878b05aaff376fc6"} {"id": "train_048967", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "66fb519495fceb742882274d7dfff0ef2d65b09f"} {"id": "train_048968", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "27a20a62c0d525016bbd8623edd096aea850d4de"} {"id": "train_048969", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "9a9501cfe33e56c5ebe5aca013e48b1d2755ede6"} {"id": "train_048970", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "C#", "moe_experts": ["security", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "757e6bd621b77b55e1ae9b683def78fb8b904514"} {"id": "train_048971", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["data", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "39eb8e0e09eec0761a08af8c4caf27cde59e157d"} {"id": "train_048972", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["security", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "17f453751e70292044a834b170f0a5cb26ce6b7f"} {"id": "train_048973", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "7d8d16c5b36be0e3f3d810a58753d504806b2ed7"} {"id": "train_048974", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "635e3d6f99728ffc04d43ba5ab9f4752536979a1"} {"id": "train_048975", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "2ce4acd63e113825f2f80f4646f57e3b72669e5a"} {"id": "train_048976", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "4dd935adcfc25b30da7283b04df12dc1f7e8c407"} {"id": "train_048977", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "e01778cc7af5e8c99b653043851d08e8d986ffba"} {"id": "train_048978", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["governance", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "2a372a13bf0bc4b08fa6a9bfb510c1cbb10b6a2f"} {"id": "train_048979", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "85aa62301756b7940eb650bcd43ea521a85171e5"} {"id": "train_048980", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["governance", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "ee4b99e496711962a25314f4bf4544d930c7c2b1"} {"id": "train_048981", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["agent", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "5eab0387d25cdf678f3a5c6d79a75cb80aedd34a"} {"id": "train_048982", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["agent", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "a7042edfd293ba84a7e31c6957834085a1f2a72b"} {"id": "train_048983", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "dab9388e774b9576cc2460402a4115c37c1ff3c4"} {"id": "train_048984", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["data", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "b8caec23098c640eeb6667c963a5d2a3db77096b"} {"id": "train_048985", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["coding", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "486e904006a9bbc83311ba8fd6db63da4b9600cf"} {"id": "train_048986", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["data", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "30935cfc7ce847c791927a578f1966a4637689c1"} {"id": "train_048987", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["data", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "e725c96b376c7af6956d9679968866f4f961f026"} {"id": "train_048988", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "7add556a8c542a476fbba624f38ef4dae5443746"} {"id": "train_048989", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["security", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "52c58f89bb4a286c4f36b126bcbf3027bb01dac6"} {"id": "train_048990", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["research", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "f62f54e96c37d1365f3d434396ccd21e9c2b0cda"} {"id": "train_048991", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "73cd8ba1321b69c7013d115fa7470aad3d2ba186"} {"id": "train_048992", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["data", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "9004dd994be2c04e08d128fc1d7233eeccd775ba"} {"id": "train_048993", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["security", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "dfe4452298469a29c2ca2fb7d0655605afbfad73"} {"id": "train_048994", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "26b97e07197070440750377051b2dd945b89f2d4"} {"id": "train_048995", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["data", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "0f7d118edb8aade090f327b2d007fcbcf3bd4f6e"} {"id": "train_048996", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Rust", "moe_experts": ["data", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "9ac57b0b5fc4e0c54ddafd207647ae7dd392329e"} {"id": "train_048997", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["evaluation", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "a7e40c678eb26e0027006b1aca7b761c0eb5d073"} {"id": "train_048998", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "c1465ca9dccd88798d53f72e45380ca3b0ecc673"} {"id": "train_048999", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "d4a96fe9b47331240f0d17415995dbbec19ae8a6"} {"id": "train_049000", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n cost = 0\n for _ in range(budget):\n p = plan(issue)\n patch = edit(issue, p)\n ok, _ = test(patch)\n cost += 1\n if ok:\n return patch, cost\n raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["security", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "dfa184f2b2c9f518c0d9a1e7bfddea1481756bdf"} {"id": "train_049001", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "01fd8a1615ad51967aefd670aa6ee5592f46b1c3"} {"id": "train_049002", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["research", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "818c8576820706624e82ebf30a6b5ca4632d65d3"} {"id": "train_049003", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["coding", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "a40931b82e283e90bee6a6553441ca75588b6bd4"} {"id": "train_049004", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "ab708636a7b1bf09de7ce2f2fa7ae1b5180b79da"} {"id": "train_049005", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["coding", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "18afabef105d2e09dc848349a04c8eb02538e3bb"} {"id": "train_049006", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "3e413bc874760cc9358a29aa347f89847e7a7a0e"} {"id": "train_049007", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "4ae3b4d05730593f293bb74b0d4ec93e11bd930f"} {"id": "train_049008", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "42fec500a899054d53d3d0df668a63bece2b1a8b"} {"id": "train_049009", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["data", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "5538edea2d380aae9aa9917a50048d02b3e031fb"} {"id": "train_049010", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["agent", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "5013d04ee9f95ab9be24b6b762d5e43ec777993d"} {"id": "train_049011", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "fff46a187bffc8a1b623f140c8d91b4873b18e96"} {"id": "train_049012", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["performance", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "d7f29584f71097a335b5a9465c4fd89157a5fcf5"} {"id": "train_049013", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "030a82835f3535a12d1e444c369db84da74c52f1"} {"id": "train_049014", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "7f7c09e70b410f221efb32aaed403677dff3a972"} {"id": "train_049015", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["evaluation", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "aa10bfd5fcc7404890b3d88500d1246e6c8d726f"} {"id": "train_049016", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["security", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "2c1db5c82fd07554fdee071d98079c83fc844451"} {"id": "train_049017", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "3384987e2cf7dbbaecf2c38bb054d08a7567d5ec"} {"id": "train_049018", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "0413f19ee9e35265b697c7d2e6e622a5623db24e"} {"id": "train_049019", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "da8bcc1883d19a730360ec3df08e3c71b2cff5c0"} {"id": "train_049020", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "ee575fed6e7e1ebceeb92059d12a65a866223d83"} {"id": "train_049021", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Java", "moe_experts": ["security", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "93b341d3ea3406b577868f7f474299562e9dbfa6"} {"id": "train_049022", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["research", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "015748858be2e2328336c4504bd9588e43a9cdd9"} {"id": "train_049023", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "f5ba595ba17a588d39017f76f7619e01ca760f75"} {"id": "train_049024", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["coding", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "e5d174644d0e679bcf10c93fcbacd396157ef35e"} {"id": "train_049025", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["governance", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "b9948c99e40bcbdb8b3eea8a51858ec52b7f923f"} {"id": "train_049026", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "12f0601f80ae3d2713b6dea7e627d43548c5a112"} {"id": "train_049027", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "c96ba243a8852f67fef5d2d6d74d8d9b6d025f20"} {"id": "train_049028", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "f8eb1851ab5e31e3be35f4dfe95bf9f942c9c786"} {"id": "train_049029", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "0dcb0763faa36c85a2aa56b409a4594097b10ea7"} {"id": "train_049030", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "fa831184e262daa2ea76e07fe36bbf7623726e51"} {"id": "train_049031", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "a665980123735676992b568e7bb08b05976c5732"} {"id": "train_049032", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["coding", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "5ad8940ff2094e8e766ec54f15cf2077db5c926f"} {"id": "train_049033", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["security", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "a962c80a73eed5aff7a0f235b1976a60814abd2d"} {"id": "train_049034", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "d28604a58e5f79626a3a6027f21c0a5365e82967"} {"id": "train_049035", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "24c1a9a818e1aa92fa89b2510138341c6ce10280"} {"id": "train_049036", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "5914414a58e64b717381cf5bb082ca464c80f34a"} {"id": "train_049037", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["security", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "fc36855214158ed2488fd6fd645de9e4b2a3d86a"} {"id": "train_049038", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "13f1cd82e94333b0e2e85d589cdd81e24e426c9e"} {"id": "train_049039", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "acebd3f42d0b18c9ddb636ebf0e1329b4548d575"} {"id": "train_049040", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["governance", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "fab1cd3119d3a15da09245dce7b835e863239429"} {"id": "train_049041", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["research", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "ae5cd47533b793bfab2b36013fc5140469e6eba1"} {"id": "train_049042", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["security", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "feefaae9b6c3d7885f36913f02ca8184a281db7b"} {"id": "train_049043", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "3feec814de6508fdb7fcbecf032a64e923cc96ea"} {"id": "train_049044", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["agent", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "627b96c9fe341cd9edd8565de3ad918ed5dbfbd5"} {"id": "train_049045", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["agent", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "e4b7bab39961e61d29e2539b60367e9e07073fae"} {"id": "train_049046", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["performance", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "3bc5b9cfe34ec492a5c4fcf6dd5ed79ec2fe1437"} {"id": "train_049047", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["security", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "4421186002c71102f4e7e9c20d01712d21f5f19e"} {"id": "train_049048", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "ca63c3b5a5bf7fcda5fbb642f153d2b8dca8c2a9"} {"id": "train_049049", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["data", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "60263ebaba7e398ee5751a757fa6c3d311f79d37"} {"id": "train_049050", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["data", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "9398efad2508dec954078fc940287838c59eadc3"} {"id": "train_049051", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Java", "moe_experts": ["research", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "430dc8d6f6f5891415c40849ea529a3d502aeda7"} {"id": "train_049052", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["data", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "681bb6968d589dbf780aca97a20985b503204268"} {"id": "train_049053", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Java", "moe_experts": ["performance", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "588f4546bea4cadd2de462f3763e25a7904c7294"} {"id": "train_049054", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["agent", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "fc8172e5aba6ff1afa58559d91bddcc4e1e60802"} {"id": "train_049055", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "11bcb929ba366b5c24d2917e1ce04f50fa0dc264"} {"id": "train_049056", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "23e91dd97ca412201230ece69adf651c83c29682"} {"id": "train_049057", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["agent", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "ab47c9706933abdc436780ecf5bbbb8a259c895c"} {"id": "train_049058", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "431e307ad096f4ff3ff06fa9a056b49798f4e9fb"} {"id": "train_049059", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "3f95c6cfd2650fe1dd171aeef31ac3f106c1d2ec"} {"id": "train_049060", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["performance", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "a580d5ad63720755814a515a227f9e2149f77458"} {"id": "train_049061", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["research", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "94cb3d728d9dabec10bb8326b92802dbc24f7860"} {"id": "train_049062", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["coding", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "dc10cf2ec3c5aa13a4addd4083f023dc93db1dce"} {"id": "train_049063", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "4b39935a0e27a0ef91447d50bd346f49837f3730"} {"id": "train_049064", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "431e307ad096f4ff3ff06fa9a056b49798f4e9fb"} {"id": "train_049065", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Java", "moe_experts": ["security", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "5773e1f22d2719f6ce9e61f8808ddac24ac51847"} {"id": "train_049066", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["security", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "8fb3aae6fa2ea3cb52d3d8825a4584dbc3f2bcea"} {"id": "train_049067", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["coding", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "0500a8687c75c74d99b88db7fff6168409162504"} {"id": "train_049068", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["agent", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "15e25afa907245dc100bfa6f77c59668178ee4b4"} {"id": "train_049069", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "3ae1403d3751381ffb898850f9165189357e064d"} {"id": "train_049070", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["security", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "59eeee43cd125751944dd69d372e0192471e1b86"} {"id": "train_049071", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Python", "moe_experts": ["research", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "f0f41b3768f6a42f410fa378215543eade9b5b7c"} {"id": "train_049072", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "SQL", "moe_experts": ["security", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "6c804547d12284c0df859cdc3508fa022a154de8"} {"id": "train_049073", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "6f7bed9caa7850e6d0135a6247e31c482758e70e"} {"id": "train_049074", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "024082c106f535de9e0157a852cc9f5b76165499"} {"id": "train_049075", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "827ac587ec6d47e1d4c56de18efe270c11b34290"} {"id": "train_049076", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "b7ca3cf22067d977c352acb6f837c8259811bdbb"} {"id": "train_049077", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Bash", "moe_experts": ["data", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "dd9546b0e846b22777430cc559372afba2372e66"} {"id": "train_049078", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "78ebc25cec94c8e2d5a28dfa3a43a29a2832f7dd"} {"id": "train_049079", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["performance", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "03977ec0e2e52efbdebbdc7de030187bc9786bad"} {"id": "train_049080", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 9,\n \"confidence\": 0.9,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["research", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "73d5b317fd2d2a08d8cadc6b9b2178f23e102802"} {"id": "train_049081", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "2f74b9ce9dacee3eb1b9937a2e64b5021a01448d"} {"id": "train_049082", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["agent", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "0039a6f98bb725bbc0f0c073e8ca8c742f3f02b6"} {"id": "train_049083", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Python", "moe_experts": ["data", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "8ca683f8d11247b9f39e7670c6549dabd12a7130"} {"id": "train_049084", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["evaluation", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "4fe8c5ff266a3d407622d1b947dc66bd7d3253a0"} {"id": "train_049085", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "92145bd696130031f4d7029bf9be0caf9c95c3bf"} {"id": "train_049086", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["agent", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "14493869096ddc2694f928fd815e211756724b04"} {"id": "train_049087", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "04dd55b72db475ef694ca860df7326441a89e221"} {"id": "train_049088", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["security", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "25dbe206cd1ff608c33ecf563e7f0783d50f7c3f"} {"id": "train_049089", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["security", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "e135ebede4d6c1cdc4773a23812702ea5bcbc8ab"} {"id": "train_049090", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["research", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "9c246f604cb8f5dc20a45b2527d135d798ade922"} {"id": "train_049091", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "46d711bb8bfbfe875e7e7fe85e03998723aa0bd2"} {"id": "train_049092", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "c40a5689bd2eb7e3402bf13878698ab2fe73dccc"} {"id": "train_049093", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["agent", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "67a140477c356c9318053545ed901b8e0e4a86e3"} {"id": "train_049094", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "SQL", "moe_experts": ["governance", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "8971790fdb0324789364a7ad5700c24ad6f1df91"} {"id": "train_049095", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "6427feeabae03f1ccedb882f4b980431369c577d"} {"id": "train_049096", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "22290eb88faef093dacd284d70570960e8b93674"} {"id": "train_049097", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["agent", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "590ee2014700ca06bac46537ff7b0ff991897b66"} {"id": "train_049098", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "adc954d11411e97725fdf91972ad670dd5996b54"} {"id": "train_049099", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["performance", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "659db7d4ea2f9a3a315ad0caca0d989a6eda40a7"} {"id": "train_049100", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "SQL", "moe_experts": ["data", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "aca413a18faa722b2b8beff812b1758572b4c2b7"} {"id": "train_049101", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["data", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "310e6abb902124f070b5e8560049c910e21fc023"} {"id": "train_049102", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "ed068703c39d6ffaa3562d216f7b3df143888056"} {"id": "train_049103", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["research", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "e1fd84203a40bedf7ea6af8968b11cc44bbb426c"} {"id": "train_049104", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "8f261c80109e98f0a89799f189fea8ab0581a22d"} {"id": "train_049105", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["evaluation", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "49ac8d318fb4e5f672c5f198e43630da3c2ecf5b"} {"id": "train_049106", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "185b987cc7723820a626db33c901833dd2cabb13"} {"id": "train_049107", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "f6b224e5e50eda76162f2f9d66af3778a29305e2"} {"id": "train_049108", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["research", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "181baef492b91fa776bc82bfcdbdb728ee96eda8"} {"id": "train_049109", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["coding", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "1e2dd60dffbdec5553b1c207ced219a162a6e9e3"} {"id": "train_049110", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "138b9672552ed58305d4f4fb8f6ed66bfa92c032"} {"id": "train_049111", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Rust", "moe_experts": ["agent", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "499b0e5b52001d0b690730bccb05c1be9d5245fd"} {"id": "train_049112", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "12efbf57256049e976b1bb92f0a0c19525aa7d92"} {"id": "train_049113", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "435f5d2993f4c31e360c19155e82d4d072a3eb11"} {"id": "train_049114", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["coding", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "ee595a3c76516ff1099632ff07bb6b1dfcb10fe3"} {"id": "train_049115", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "df514ccd4c0d3a528f576112a6dc4e8f3c9f1c6c"} {"id": "train_049116", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "54b37f3f26a1782f822316f4243d8e0d12079f5a"} {"id": "train_049117", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["governance", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "19f19c1fb746aa966385e88d834b2984854dbf85"} {"id": "train_049118", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Python", "moe_experts": ["coding", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "9f0677bcb554a76ce8bd8c03421b653036a6f7c9"} {"id": "train_049119", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "a2d0565068fa1cc53d755abea691ce478c176664"} {"id": "train_049120", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "e3883fac3bb79bf2748baa5e53c720435bc631f4"} {"id": "train_049121", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["governance", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "6c3245eeb9e14d48590c62b5a9b222f38a03b094"} {"id": "train_049122", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "afec2397e66584a581d8d4fe1dc97a8492d690d2"} {"id": "train_049123", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "SQL", "moe_experts": ["performance", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "c04ba525844d1a45e6f50a66b193129e64aa7e0c"} {"id": "train_049124", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["governance", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "23b22d94d350d4a99b6f1057537814272294bdce"} {"id": "train_049125", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "941222a16fca0c243eb92a2173ac01375586a930"} {"id": "train_049126", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "cc5a2c5acb83b7aef05158faa7aa563b20710c14"} {"id": "train_049127", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "14dc1b180d4e7eb3cb7af540066b7fbd36efa916"} {"id": "train_049128", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["data", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "302ef3bfe76efc1ad67a3886cab9caa5dd0c3dc6"} {"id": "train_049129", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "4f23aa62fe6420adddae9f95c8531d68a738e0c2"} {"id": "train_049130", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "cfbca6aeaf20f5cbfa819f8e701718b53b523830"} {"id": "train_049131", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["performance", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "1f78697c18566e12753469554cb542e747a7e674"} {"id": "train_049132", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "da8bcc1883d19a730360ec3df08e3c71b2cff5c0"} {"id": "train_049133", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["coding", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "99a1c15b224927b16de0d589e9d420d201e7e0b9"} {"id": "train_049134", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "28f3c9efad2040432f27573dd3b2b8e0a899692b"} {"id": "train_049135", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["performance", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "884bf890f028ac74011d2cbc3623cc7962b11fb1"} {"id": "train_049136", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "62cf44faa50d5daa79fce65485039fde428d45dd"} {"id": "train_049137", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "fb4f38c8d648888930c21f8a66ceb629e320fa36"} {"id": "train_049138", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n cost = 0\n for _ in range(budget):\n p = plan(issue)\n patch = edit(issue, p)\n ok, _ = test(patch)\n cost += 1\n if ok:\n return patch, cost\n raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["evaluation", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "c3daa895800341a04625ad94844e211d9b88ac84"} {"id": "train_049139", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "fc334d52f193bac18c35da761c59165ee2af6693"} {"id": "train_049140", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "aa4ea851545b3039d6cb7accb33e210fc30d72bd"} {"id": "train_049141", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "075574f79581bd67d4e2c059876e872898fdaf33"} {"id": "train_049142", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["performance", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "e0834b968295a6b82fc49d2da7c5f977a3d40a75"} {"id": "train_049143", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["research", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "1c25b14cae089269b8849fe310c88e32059e4281"} {"id": "train_049144", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["governance", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "790282a4f09646b5aba2eb49b6846280f5a8e3c8"} {"id": "train_049145", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["performance", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "f5829d6f3560219625772eeed4f0c57de2724908"} {"id": "train_049146", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["data", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "534688190abb99d54e6e3fb13d21be4ae7c47424"} {"id": "train_049147", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["security", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "e93515118628110d6e13a63106ce8615b1002d5f"} {"id": "train_049148", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "3849969f940c0fa51567f217d291c752549df16c"} {"id": "train_049149", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["security", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "5849c746b270dc4f90f747696f5b2b6d33ceab21"} {"id": "train_049150", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["performance", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "d43387e6ef70a660f45a0cc449cff579c035f969"} {"id": "train_049151", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["security", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "144a7b9be20bdad3301344db78afd82b5a028234"} {"id": "train_049152", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "e2ef2254d0de76502a64ee8f2a3d4c6bd3ea73be"} {"id": "train_049153", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "1ebdbf509b733aa22349026fdeaba9d939298494"} {"id": "train_049154", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["security", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "07348b8dd0d4d0f07b5c953cc88772531f63a37d"} {"id": "train_049155", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "ccca1c2eba7ab99baa1b40a56d5fb51096769cb6"} {"id": "train_049156", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["coding", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "327f54c4ed82a9cb593e4ece5bc0841212e605c5"} {"id": "train_049157", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["evaluation", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "26b38ac221be818cdb85862949192f5b417dadfb"} {"id": "train_049158", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Go", "moe_experts": ["governance", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "6c7f638d71d9144c251f2d4451358f5f6da20a38"} {"id": "train_049159", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["governance", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "8744ee28c0d37ab569c48323b7aebba68102b32c"} {"id": "train_049160", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["data", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "20e8766b94a34efac577dbfda7f8fd0aad80ebf9"} {"id": "train_049161", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["governance", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "a3d8eea8034f2e5d791c7b774576faa609002562"} {"id": "train_049162", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "ac289dc75f73bfda237cd466378c808da4010546"} {"id": "train_049163", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "4a1eb54c602015619f07d4397fa639978863546d"} {"id": "train_049164", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "964c2c3bc25a751327c15380010a1303a7834ad5"} {"id": "train_049165", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["agent", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "c81b815f6f769fdc70d3494975076d7dc231daf5"} {"id": "train_049166", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["research", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "af7b1fb5bd91a120eb56b675896a5a28d3a8b35f"} {"id": "train_049167", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["security", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "571deda37388ad5495c6947f5b254c8a282c1037"} {"id": "train_049168", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["coding", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "f835c21d26ae1596875d28a4924b025090785cde"} {"id": "train_049169", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "b92ee3ded19592c685a924b93d40bc3df8853399"} {"id": "train_049170", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "12e5bb9109bd900d149323062de445e73c87b444"} {"id": "train_049171", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "641cff9996fb4ab584b8b60c7aec518e746ed5be"} {"id": "train_049172", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["governance", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "688307756e278c3e7b782115d4796db11b68705f"} {"id": "train_049173", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["evaluation", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "a6280dfc2367a1ca22aee7b9052f15178637045b"} {"id": "train_049174", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["performance", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "d7fde81295757c61ad080d58f4b57abc5dbfb575"} {"id": "train_049175", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["data", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "db5e9ba8f950e6ede050ca5e3ecc76842e690572"} {"id": "train_049176", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["data", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "197856f9e730ef98b10faa42255cffcbc2a3819d"} {"id": "train_049177", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["agent", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "a14669b41726d6974a366b69874c60fe5e8cbbc0"} {"id": "train_049178", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["security", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "e75a9755a29ef0b0c1f81fdb8bbcb99cf7edb40a"} {"id": "train_049179", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["evaluation", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "ab6114b7d8600d4770fbe0a44d273041347f5be3"} {"id": "train_049180", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 9,\n \"confidence\": 0.9,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["evaluation", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "ded7e22e5a678409f69307a5df3eba25e9279007"} {"id": "train_049181", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["data", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "c3868ccdff7bd3af171a3ee24df60aabd2d346a6"} {"id": "train_049182", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "4ae776ad8e5b652da92f0b9ff69ecf961d4d92fd"} {"id": "train_049183", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "b9e46799f8a183f17a489612745b0c61d3623b20"} {"id": "train_049184", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["data", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "4f1408d365ea118be3b52055f65f377ec94699b7"} {"id": "train_049185", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "348fe4e0e893334d269e3e4c9adf647b24814766"} {"id": "train_049186", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "806633ab3696447bd21ff02ef2fc12d7d25bf85b"} {"id": "train_049187", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["governance", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "e55896dc9aa4e86d51b0a092cc1df5ca0bf1af55"} {"id": "train_049188", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["performance", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "c680235903a664e96495f36a169bed14b2cf9e71"} {"id": "train_049189", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "5349c12ea750f9f5437883b5007b8e6422faf11c"} {"id": "train_049190", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 9,\n \"confidence\": 0.9,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["performance", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "0b67300275d1e2ee18221440f803dd31ae4f5e5b"} {"id": "train_049191", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["data", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "07bd75c93a6e9c11bf0c8cc2f7d863d0712d23d8"} {"id": "train_049192", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["security", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "157a42e04e3833f6cb23adc3190c8ddc67cb12e1"} {"id": "train_049193", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["research", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "a1e91fb57456720621b799cc80b6c4ef58784f47"} {"id": "train_049194", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "135804d415b41200a7a862bf6f40c612485a8dd7"} {"id": "train_049195", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["coding", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "28272db1a70b1ff616561e4dee0fac2756d55736"} {"id": "train_049196", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["coding", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "6cddfc7835daac641742f6adcea41786b4b11952"} {"id": "train_049197", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["security", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "6c5a81c0b4409e1fd77355d1345e96ef2b374b94"} {"id": "train_049198", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["performance", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "16181c4d6932336e1d46805e501d8021e92634ad"} {"id": "train_049199", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "fbb831d1c9d9cece6ddb83923c6b7f690d489708"} {"id": "train_049200", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["performance", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "7a6cf89ab6df6887d9cc4d91247863e7ff6ad01a"} {"id": "train_049201", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "3eff1572354ea78aa37096a6825d7c488eedba0f"} {"id": "train_049202", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Java", "moe_experts": ["research", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "759e49046bc54c2f995abc79ad81bfd6baf0db96"} {"id": "train_049203", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "080140e317e42a8bdda7316197d11a7091d6d45c"} {"id": "train_049204", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["data", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "bbb76155c1d52f920805a381ff906096f7937fef"} {"id": "train_049205", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["security", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "270ee5222f0c69059e40df7074cab251dc92c11e"} {"id": "train_049206", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["security", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "dc95b3d4b216b7f25371614dff49088c869edccc"} {"id": "train_049207", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "97fce67ceba269c0c18bb537acddb6efafb8c924"} {"id": "train_049208", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["agent", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "ebcb3cd8f7f994fc60f0251892abda2ab3049400"} {"id": "train_049209", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "8f35e25ca9cbe6722837ebe3efe9627915305301"} {"id": "train_049210", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["performance", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "14c39db7a3cd0b7990ff10f4e5c7c3ebf5dcb4f2"} {"id": "train_049211", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["evaluation", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "05ac77711c79499ae41fb177f1d018a13563fa42"} {"id": "train_049212", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["performance", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "b4d46133b846c2125626c0d0057ee890d65233fa"} {"id": "train_049213", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "4f473473844940aae289afbb164a73f594277027"} {"id": "train_049214", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "9b67906591ac381f396242ece7b1f9e7353050f4"} {"id": "train_049215", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["security", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "cd038faf2135e99eb3a0ec40a68ecfebdb8cd28a"} {"id": "train_049216", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "bf4117d0f42a525689c1ab5208720ce255d6a4ee"} {"id": "train_049217", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "eac3da1aad123f68ef96666a5c5055b450218f39"} {"id": "train_049218", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["coding", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "1700b178954f99f97a40c0f0d0f5390934b97312"} {"id": "train_049219", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["performance", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "e09169b21a81bd641f47799edbd8c75d7adde409"} {"id": "train_049220", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "28d4ad5b8ddd9cbeaab38fc8dd09a2feaa065ad3"} {"id": "train_049221", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "0c4678ee7ab3a6169fda579309172f9ad6d66824"} {"id": "train_049222", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Bash", "moe_experts": ["research", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "9fa5e5e4ad3b025a4ddae635ad59cc7ab1cd6de8"} {"id": "train_049223", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["data", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "0f6695196ab21acc8ebb25f4111f2e6e674473ec"} {"id": "train_049224", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["research", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "91d51f9cd9074963e93834b8fd75e331df164aff"} {"id": "train_049225", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["research", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "738f2ad90c5ccdef9947014b349b23aba521d52f"} {"id": "train_049226", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["research", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "35107cdda39bf4acb2a98cba477434f606cd4377"} {"id": "train_049227", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["performance", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "e5af702e19a25c51e9edea4db8cc21202133d30b"} {"id": "train_049228", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "4112a719f6d30822e304ec3f5ade8bdf49a822b1"} {"id": "train_049229", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "65d996ce87d918309595e6aeca3ce88a3750ca98"} {"id": "train_049230", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "SQL", "moe_experts": ["coding", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "0f2b04cdbe7e2d261cb831474d0a1124736c8d99"} {"id": "train_049231", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "daab0be652cf623b61f618c17d73ec5d63e0d8eb"} {"id": "train_049232", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["security", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "25d3678a2eb725c2d1460164b6d399078e5551dc"} {"id": "train_049233", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["agent", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "feefaae9b6c3d7885f36913f02ca8184a281db7b"} {"id": "train_049234", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "194d2e90a513ce082cde373d29f8057b0ef0cf04"} {"id": "train_049235", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["performance", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "d31d1b7a63f1c4bcc64ee2a4d8162a6938473a77"} {"id": "train_049236", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["research", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "d77bee4bf988cb2694ddd445fb71b192aa2936e9"} {"id": "train_049237", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "96ab87e330ba7454aab8870537676c5ed71bf411"} {"id": "train_049238", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["security", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "2c9750148f8c60a8ca8def905d7b7fcdb61f3d3a"} {"id": "train_049239", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["coding", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "378126deef11d8b63402d63a695ef78b27f1573f"} {"id": "train_049240", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["evaluation", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "2eae75e7b4edba4ea9c09e389e71f0bc48df1eeb"} {"id": "train_049241", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["coding", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "f9a39b3f7e966cb0ac83d5f0fb7f62fd07964b21"} {"id": "train_049242", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "2f386dedee7d5070a112e01f22cce949f36841d8"} {"id": "train_049243", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["performance", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "591238f4af1a84b64c8615d2176945ecee268b15"} {"id": "train_049244", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "b8950f6ad0c8903f0ced293a24028f927969738d"} {"id": "train_049245", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["evaluation", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "7b8df5e6caab5e1a62fd2c901abf3b336134bd9a"} {"id": "train_049246", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "dc5d9d9a0e89a7f3e7891928275ef1d3cd212502"} {"id": "train_049247", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["research", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "7e46d3934818a3a1e8cf939986f1c5bf7704a0e4"} {"id": "train_049248", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "fc36369c85f3e8cffedb2cf2eaecd1f66eab546d"} {"id": "train_049249", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n cost = 0\n for _ in range(budget):\n p = plan(issue)\n patch = edit(issue, p)\n ok, _ = test(patch)\n cost += 1\n if ok:\n return patch, cost\n raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["coding", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "dfa184f2b2c9f518c0d9a1e7bfddea1481756bdf"} {"id": "train_049250", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "185abdc1c632ec948f6508b16b10083358ec81ae"} {"id": "train_049251", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["research", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "5cc20219451136e25e60d83a7132284643faf250"} {"id": "train_049252", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["evaluation", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "ef9cc0cffd531b25ddfb41b2a18b9d9b5acc5c31"} {"id": "train_049253", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "SQL", "moe_experts": ["performance", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "0f865e5a5e433ead8bbe34f561ee277c659209c6"} {"id": "train_049254", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["performance", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "9aa274526ef46fce38e897aa35f6f260e325491e"} {"id": "train_049255", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "084c31b331cc2942dc5a46a4b8f065e795ac892a"} {"id": "train_049256", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "de8ea03978c26e95489e71e404dc743e6f7beb96"} {"id": "train_049257", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["performance", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "c26fab5a069aaecda2fbf5ac5e97e164fa54dc36"} {"id": "train_049258", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "97d9dca1088e97a7dcc7714de3f3048a0300f0d5"} {"id": "train_049259", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["agent", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "2772d223a63a7da4a1a4cbde6efa7b8aeb5374c3"} {"id": "train_049260", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "SQL", "moe_experts": ["governance", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "134939947da988224fb61bfe172145afdde90a2c"} {"id": "train_049261", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "0a58a9dd92c407cb9aeded875896969443396f7a"} {"id": "train_049262", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["coding", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "d20fc3392dffbcf112b63ef6029b6fe0e6eb7055"} {"id": "train_049263", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["governance", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "a45de7f9867b5afd1482c8dde49d874fd01e73bd"} {"id": "train_049264", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["data", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "a9be2e25a1c22e35ef5ce123d22b386a14d14769"} {"id": "train_049265", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["agent", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "77e2e402c1d3dfaa5a824daa4cbd15ebadad75ce"} {"id": "train_049266", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["agent", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "7b0f53cf307ce627c441ebe63668ec2b92185db7"} {"id": "train_049267", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["evaluation", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "029a95928ba068307ac5c3fdd1edccd23dca37cc"} {"id": "train_049268", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "7f5fe1b1be78eeaa3d192ba893018c64a631941f"} {"id": "train_049269", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["research", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "c42b8424f241f9f4286896b4c9d9fd47a849f344"} {"id": "train_049270", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["coding", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "91b51738183f1022d4dd902317752a9179966705"} {"id": "train_049271", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "ace0459307faf1c7933f1bc61a4d8bebfec67088"} {"id": "train_049272", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["performance", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "0ef83ff30527ab8fe4dd167cf6eb69321e5c0662"} {"id": "train_049273", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["agent", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "554787c2558320679f6d042ed450e985a9332d7d"} {"id": "train_049274", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["research", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "1fd979b03800940acd0255f84aaab853b24f8513"} {"id": "train_049275", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "5756b757b244cef33ecdec1ee8e5adc8f555885e"} {"id": "train_049276", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "e3d723e8707a381edd462b7200cf2930aa8c86ff"} {"id": "train_049277", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "78ee7058c1e51ec000a1388aefe46c00eb0aae9a"} {"id": "train_049278", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "b7499897ba4c13bc968ecb1c55fd0dc298b35fd7"} {"id": "train_049279", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "a9edfbeca8ee9dd9e87bb4c7872dd8182409244e"} {"id": "train_049280", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "53c35dc0bb19eef8f3dc8ef0ffad89c09356296b"} {"id": "train_049281", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "3e6ff4276dcf3dc340be54f876a88273be8a4def"} {"id": "train_049282", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["governance", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "00c251b0490fd79e2ac5b9093a088cfff9c4108a"} {"id": "train_049283", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "517a43b64643d1a39608715e99ae171119ffa111"} {"id": "train_049284", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Java", "moe_experts": ["evaluation", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "165047ffa0789b605c5b29cfd9eaf6dadf34ffcc"} {"id": "train_049285", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["governance", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "61a126decf60763e8d5c7208cade53b9ca6f0268"} {"id": "train_049286", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "c75adfb706db62a2ede6870fe35b70c685aa42bc"} {"id": "train_049287", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "6843973934e949620fc86fecd326f190f341f455"} {"id": "train_049288", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 9,\n \"confidence\": 0.9,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["coding", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "3ff1f9091aa282064ebc0f84c6a1cd732fea34ae"} {"id": "train_049289", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "efa409dd9f97cc64cb5ec115d62390b6b5c11f67"} {"id": "train_049290", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "4ff31c7237d2391171777f7863a0dd1ae37ca57f"} {"id": "train_049291", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["performance", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "26b52a4d137be54eb0478b24db042faba0259238"} {"id": "train_049292", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["coding", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "6b3cc378fdcec6ddbd8d2a2d20ba7e52cd9f6254"} {"id": "train_049293", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Python", "moe_experts": ["evaluation", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "ae32b9bae03142547db515d17f6d68aa63cffbeb"} {"id": "train_049294", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "5fbcfd5d4ef79cfbbd52c05e9fe716920acf2392"} {"id": "train_049295", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["governance", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "8fc47c6e3b5dc6440ef69c9bc66073febdd110c7"} {"id": "train_049296", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "5b816b046b3a4e7416a53826a049500cbcb45796"} {"id": "train_049297", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["agent", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "3676a82b5282446d0bad16da239064030cf0c56f"} {"id": "train_049298", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["data", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "e8863e49f4c7666692ac5e726f97ae862135af71"} {"id": "train_049299", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n cost = 0\n for _ in range(budget):\n p = plan(issue)\n patch = edit(issue, p)\n ok, _ = test(patch)\n cost += 1\n if ok:\n return patch, cost\n raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["evaluation", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "982312b2f1458721f9b911671d4c8d57638a807a"} {"id": "train_049300", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["evaluation", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "3d16f02808a1c2fcba6a17441b97bccde3184667"} {"id": "train_049301", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "83e7e6491137f9138d380e74ac272ef6fde5ab0f"} {"id": "train_049302", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["coding", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "3bab1620204b990af755b6bd97cbf42c106bf088"} {"id": "train_049303", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["agent", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "81ac9a92638657745501e6ed7a8c89fd017b528b"} {"id": "train_049304", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "83db23bbcfc1ef4e5d2dd37dd1fa41038c4cd8d0"} {"id": "train_049305", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["agent", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "82a1f0f7da70e08efc6b60079216f213007a0d18"} {"id": "train_049306", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["evaluation", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "26191b38278f6b9f3cd619ab335e50e4d25109df"} {"id": "train_049307", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["security", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "3b3afcb9e21f34f46c5aee759b7a4987917528a9"} {"id": "train_049308", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "f5087718a0555aba840fe9da344a86634c4fa2fb"} {"id": "train_049309", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "8c87d8abf1be9a92e49eda1e69f635c8708b2670"} {"id": "train_049310", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "5e1c833b711b93643a93ea6d44ed548a1804ff01"} {"id": "train_049311", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "484372c47e581e062761335b926b7cefc5515ee3"} {"id": "train_049312", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Python", "moe_experts": ["performance", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "3c509973f1fb3e193de9787765d6367f5f4a1680"} {"id": "train_049313", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["data", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "4cb2d692e0ce6b5b277d42ecb091a5facf172bc0"} {"id": "train_049314", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "791e7f72f39fddee7bd68a28223ecb15cf25cf79"} {"id": "train_049315", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["data", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "0b7a2ce5f783cd4fad66dc543ce6a043c1fd78ee"} {"id": "train_049316", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["data", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "6226f2e076c2927791218a5ff1cd450a4764d5fe"} {"id": "train_049317", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "33335c31c113dec10913d9b94dcbd48c122385cf"} {"id": "train_049318", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "614662976a4d4d8333b340b4b38a5f94ffec3651"} {"id": "train_049319", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "7236449ebe53b096eae2dd5a916818edb5ee05e0"} {"id": "train_049320", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["coding", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "a4ed650398cc1ba653f87617147e3648ac46d74f"} {"id": "train_049321", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "dd85ab6ee7137cb3342793095d1a89e2a9adc72f"} {"id": "train_049322", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "0c4b164cf4996d49a7ea6b1d2fbde6e20ba18c4c"} {"id": "train_049323", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "42ce8843de6175ca849eb29a67c0645bf991fb64"} {"id": "train_049324", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["security", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "096a0f9ce265dae6a9d6c68d74955b388fc320c9"} {"id": "train_049325", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "fc6b9dd11fa7ccf7ced6b508984dd37dadf0e604"} {"id": "train_049326", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["coding", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "4476809e6bbf860c673e245366d4f455967a9875"} {"id": "train_049327", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["agent", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "ee15ffd4255a04a642601c3731bae4fa64b00019"} {"id": "train_049328", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "12e3f793336f1d27990c806d5bce8c29f70d42f7"} {"id": "train_049329", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["governance", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "918c8ebdf748952eab9e471b9d1f122ae75e8bf8"} {"id": "train_049330", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["evaluation", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "e920e28478dc5def356b38c01937afbc5fdf9656"} {"id": "train_049331", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["governance", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "64264f4d3dddb49a4ab71d62305a831325df3031"} {"id": "train_049332", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "4c9d566bdb36dc307085dc481f0a7554127b628a"} {"id": "train_049333", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["agent", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "04eb598cdeba1986a4127e9c6961de4761ea6929"} {"id": "train_049334", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "fc6b9dd11fa7ccf7ced6b508984dd37dadf0e604"} {"id": "train_049335", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "02d83c4717d7125032b5f86324e9a07bb97e3159"} {"id": "train_049336", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "0d2bad129f144eadb2962dedf741704fdfeeb57e"} {"id": "train_049337", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "b18c9c6a96c5b5444ab8099edb8a4c6d7425c9a7"} {"id": "train_049338", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["agent", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "21f57f818b82078b7b9066b97207d213d98edb27"} {"id": "train_049339", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["security", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "115f6e2d775d940cd51d27c4dc37e55b71257464"} {"id": "train_049340", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "a7d1933a796cb1cc36f3228181ee9fa291d4d157"} {"id": "train_049341", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Rust", "moe_experts": ["coding", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "302c2b6340e75e54a4a7bc092ad825cf80302c43"} {"id": "train_049342", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "cfc4db1b593f48c405d0df59d5cf838c2356d702"} {"id": "train_049343", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["data", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "8e36e4c9ca529a3e667dc605dfd88c8ab0081802"} {"id": "train_049344", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Java", "moe_experts": ["data", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "87e760d5ebb13c7c9e633014615a60319d9504cc"} {"id": "train_049345", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "f6a3f1dfd77f6ee91606c4bd3b6ec13fd99b884e"} {"id": "train_049346", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "d28eebcedc2e6f75a1ae4a54708c98143c748731"} {"id": "train_049347", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "f2ec07e7750dbfb7293bea931b07177088ca9552"} {"id": "train_049348", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Python", "moe_experts": ["research", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "73954516cd7708a9fa02050a711814fce64daf12"} {"id": "train_049349", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "SQL", "moe_experts": ["agent", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "cf6feec8b15a87c5b56022e4944c49b9bc02f25a"} {"id": "train_049350", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "f2f51c8f7e077c0dac7325d699c7b3dcf5e85ace"} {"id": "train_049351", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["data", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "cbc59448eed4e75869c3b55717e9acd14aa1d611"} {"id": "train_049352", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "cf0e8525fd29816f1f9c4a9c9789f6e58ca1d686"} {"id": "train_049353", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "c875efc487bfb725b5b1f8f88234212e1c704c60"} {"id": "train_049354", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["data", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "f4ad8e7b329e44edf63be41f19fcc96c99355e70"} {"id": "train_049355", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["coding", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "18083aa6b4e3df14bc5d5eb60bb56035d92a35fd"} {"id": "train_049356", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["evaluation", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "b5294d4ff15dbeb88cf73d6dc43d04ccdef16389"} {"id": "train_049357", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "6b95698809564756b7bb769df0ed0814d601d244"} {"id": "train_049358", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "9d916793f091bc28e162b332c0ce112ff15dcf95"} {"id": "train_049359", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Go", "moe_experts": ["research", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "8c4932e7e441c98dfea975d4c73206980dc5d988"} {"id": "train_049360", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Go", "moe_experts": ["evaluation", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "31da926341d8bdf93950a3c8d48b3f45e7e254a6"} {"id": "train_049361", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "c8e8462bbbc5ca142e6ed6a99be285a59ab3b0bb"} {"id": "train_049362", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["performance", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "81dbbe0239708343e63a50d41ae9de582d932876"} {"id": "train_049363", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n cost = 0\n for _ in range(budget):\n p = plan(issue)\n patch = edit(issue, p)\n ok, _ = test(patch)\n cost += 1\n if ok:\n return patch, cost\n raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["security", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "41f203c2e54f4ae75582ac280d79380e96134a31"} {"id": "train_049364", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["performance", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "02d83c4717d7125032b5f86324e9a07bb97e3159"} {"id": "train_049365", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["research", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "8bb2fea9655fb2ea835b190c026540cda7bd61ef"} {"id": "train_049366", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "501fdb9574ced461ae8afddc7df576a1abaf981c"} {"id": "train_049367", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["governance", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "e3a8dfe81a22afbcb2edb1f43a739b32a211622c"} {"id": "train_049368", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["coding", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "52cad26c790c8a3e96c4f1a5ce94b49783c79e72"} {"id": "train_049369", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["data", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "468ac2739014d4a4d943c8c6dc8ab49b511e0714"} {"id": "train_049370", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "5d3f3e9073fb7e9178ffcb3f3b6f44c5b1b4a5ac"} {"id": "train_049371", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "1826b134c1eb4a551f47468fed6a4fed59431abe"} {"id": "train_049372", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["evaluation", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "4a48f95524ade2a2e0ead54d4b5eea0d9d2c65ee"} {"id": "train_049373", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["performance", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "3cdc405ffc4353eda05d3e210ce235de193ffc9a"} {"id": "train_049374", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "4bb915d86164fdea7c6fc906622c8e227df94be1"} {"id": "train_049375", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "2c52efcb8d0be30e5499aa35762448f5a0c9b3f8"} {"id": "train_049376", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["governance", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "e0ff3a7a273ceeb883099a9701245cedfaad9176"} {"id": "train_049377", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["security", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "028bbfd8a485a0d3b905582ad7724966a0cd91e3"} {"id": "train_049378", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["research", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "1f1ca4a63fbffd4f012f3c3d0b41ae2bdfda46f6"} {"id": "train_049379", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["agent", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "3124eb984ba349ad49719d6beea7f4874befe4d1"} {"id": "train_049380", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "C#", "moe_experts": ["data", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "ebcc4227996926853a7f878b4554a5962ff32d76"} {"id": "train_049381", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "339370724fb09fa98eca58ba2dfe55483179d2c6"} {"id": "train_049382", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "26e65aff95d5c10c38c1f658a460295577b68994"} {"id": "train_049383", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "65146c364aac84c845a7576dd439674933146876"} {"id": "train_049384", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["evaluation", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "7e46d3934818a3a1e8cf939986f1c5bf7704a0e4"} {"id": "train_049385", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "438dabaa2dde99ab9225ac826816d34955ea6099"} {"id": "train_049386", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "9b2415b173755d0e30b9398f5b9657f236b77e97"} {"id": "train_049387", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["research", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "471557ee00b972485fed12165d7e3009964c37ee"} {"id": "train_049388", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "c343390438a6beab0890526f1602f35e96ca46b6"} {"id": "train_049389", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "b056c3de31a55e804e32a9ef1d62c966f64fcba1"} {"id": "train_049390", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "b7b8f0372f990b8941209012a03a96776d1139b7"} {"id": "train_049391", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "916f8895aa83c5ebcfed036411ddc8af76ccfe20"} {"id": "train_049392", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "01f845fd4f04eaeac6b2f0b167d48896e3063f64"} {"id": "train_049393", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "C#", "moe_experts": ["governance", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "cc24798f77ff93deffbf1cd6658435e7fad2b16d"} {"id": "train_049394", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Python", "moe_experts": ["evaluation", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "b8d4f02bc144cce05078098aaaf4d456f9eb0784"} {"id": "train_049395", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "aab8eda2415f049e65f2f08339e45cda998c35a0"} {"id": "train_049396", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "5c1f61cf0e63510899984de101720f2d47f69a4c"} {"id": "train_049397", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "9bf91afb97631df08c03ff339dc3e32d4852f17b"} {"id": "train_049398", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["governance", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "7c65efebc8f02b49847e44aec0fe3ad854c04744"} {"id": "train_049399", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "7c360b73844c9cb3841e3e2683d51f5c5bd81c00"} {"id": "train_049400", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["governance", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "632b4222bf0421574e4ecb59f9f417a1acdf3647"} {"id": "train_049401", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["agent", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "86056c4caa7412abb7246293b985975235be18cf"} {"id": "train_049402", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["governance", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "e1a2e3b20fba4d6d2490067283360efbacf1736f"} {"id": "train_049403", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "44b4818a43dabd6c905b6053a934f2c1e50a2caf"} {"id": "train_049404", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Bash", "moe_experts": ["coding", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "079047e1ab7bc10ee7be630eb47cdfa4e7b59b94"} {"id": "train_049405", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["performance", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "769df500987ee1a3d7f22c466b3eec9cf4df451d"} {"id": "train_049406", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "3dc19e407e69f8ebddaf849a86c8f2fc1e59b2f0"} {"id": "train_049407", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["security", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "17e6f90bd1e38b9a6bff00b22a6a518e0b13d03f"} {"id": "train_049408", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "c26ac81d188b5f5cbd53349cc46f1bdc26cc1b21"} {"id": "train_049409", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "4a88319fc18ed129de3a0c266537e1799d0a16ed"} {"id": "train_049410", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "66f51f2946c149779630259b17e051877b7cd4e5"} {"id": "train_049411", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["agent", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "d74e06325f69ab7c7ea41d81be91aec5c80bb176"} {"id": "train_049412", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["evaluation", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "470761b99ef5eca503e6b06afca5ef483bfcd070"} {"id": "train_049413", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["agent", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "025ebb27ac2fd8372f0821ef75d1d100b5fc172b"} {"id": "train_049414", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["security", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "85af969ccbc0ab2a6b0381dc9d7f244f56af1e39"} {"id": "train_049415", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["security", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "eedf4b19c30cf437af45fb01fca4531357c0ffdb"} {"id": "train_049416", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "58bc280d3a8d8182f86aaadc40c13e593a7df6a9"} {"id": "train_049417", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "016ee965bf128ab58386bd98e14d6761f58de5e7"} {"id": "train_049418", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "1a5d786a4aea51f905714e2101746c8549cf6f35"} {"id": "train_049419", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "8712adbf79c541468c2912e2a591eafb60609076"} {"id": "train_049420", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["data", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "b682703ac19ca74c9009eac8a2e05731d418b111"} {"id": "train_049421", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Go", "moe_experts": ["research", "governance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "02c58c866afab0b9b70edc846d6d4a6c5e641289"} {"id": "train_049422", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Go", "moe_experts": ["evaluation", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "32ba8d274825938c0083bc7f44e6cc806b694539"} {"id": "train_049423", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["security", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "a6ae2c946cc9d41b7cbc8d7681d302556546a366"} {"id": "train_049424", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 9,\n \"confidence\": 0.9,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "eb4eb1d7b8bc7b43e4bd094ad4af52e3bed2208d"} {"id": "train_049425", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["security", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "e649938b8e9ac8c2d0266093800028ee617012b1"} {"id": "train_049426", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "92518a3a15c6f240543a48497cf7d182bd799b7b"} {"id": "train_049427", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["data", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "3586184280b78b6d2788e2c9c246c090852fc0ac"} {"id": "train_049428", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["evaluation", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "fcb993d46e724c83d5df613ca81c912225043ab8"} {"id": "train_049429", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["data", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "07ef8d2cf2e4b29e323892fbb7c1e51bacb7b123"} {"id": "train_049430", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "SQL", "moe_experts": ["data", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "c012639704990f535e8727240b5876451b403b38"} {"id": "train_049431", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["performance", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "a64f2796365a21052c264ff225f031202cc5dab1"} {"id": "train_049432", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["agent", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "ee7e4de174ac5676892f115e0f15e1c761c2b356"} {"id": "train_049433", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "019cfde6172121ecf696dbb86c555a4f191b6d2c"} {"id": "train_049434", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "9e2f2ccd832c64451106863601c25c452dc9a13f"} {"id": "train_049435", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "fda43ba2aef596e19ac241132fd3ac1e33e6da49"} {"id": "train_049436", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["research", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "20ba8fda0e2931928b79254ad08ef3fe45e0f4fd"} {"id": "train_049437", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "388929d7d5a69d45ff0ef9609b6e51d36352b27d"} {"id": "train_049438", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["security", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "b8e0c259ffa6dae9076806ed35691c3435fa2b7e"} {"id": "train_049439", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["governance", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "b10157c4f63f5801357cc6721c65bb331d552546"} {"id": "train_049440", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["coding", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "deafccee74d1b9fb79e9d6937bbbd4eee9d9f821"} {"id": "train_049441", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["governance", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "6cea7d758d2b223dbec508b61c056cb40023f5b6"} {"id": "train_049442", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "8571463c4071e7a885e85afdb2badc15de5d722d"} {"id": "train_049443", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "b056c3de31a55e804e32a9ef1d62c966f64fcba1"} {"id": "train_049444", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "fb761a84a1602a18f9fac48d5257b6b484b7e3b6"} {"id": "train_049445", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["data", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "0ad540b12966c20248420eb9e63082f25ff3b843"} {"id": "train_049446", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["data", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "979ac33282efde343460ec6295a0d17b2d2181c5"} {"id": "train_049447", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "8cdc1315fd94fa04a679cb34f45bb9f7df8f19ac"} {"id": "train_049448", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "c6e8616d79573bddac1e224cc84dd1a513f1a749"} {"id": "train_049449", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["evaluation", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "38dac40732734c769e0e715fa457f43ce70dc1bd"} {"id": "train_049450", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Java", "moe_experts": ["data", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "0c02f1a0050e08c50be0a050e91ab4419b576434"} {"id": "train_049451", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["performance", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "d6a51c125098d5cdba72fd5a539e755098cfef8d"} {"id": "train_049452", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["data", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "01e9760e4cc11ca1a4fc24488a7443aedbf5482d"} {"id": "train_049453", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n cost = 0\n for _ in range(budget):\n p = plan(issue)\n patch = edit(issue, p)\n ok, _ = test(patch)\n cost += 1\n if ok:\n return patch, cost\n raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["research", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "b790fee69be4ec405376385cf588108f84072725"} {"id": "train_049454", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "897f9657b5e5f7af94d197755ba53a68c8bec7b2"} {"id": "train_049455", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Bash", "moe_experts": ["governance", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "a595aab6d7477150b4d73e09e7daf007dec595a7"} {"id": "train_049456", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["agent", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "a2f2f5a0aeb3c3b9843c23a0ce1f2679f990068d"} {"id": "train_049457", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["security", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "796da9d6ce7bb07dcfb44e267062f6ecbe08c76a"} {"id": "train_049458", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "a6e6bb42214aba0e44edfd532c420b686f2ba908"} {"id": "train_049459", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["security", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "c86421fe01abfc849d8dd6f316e881350a143903"} {"id": "train_049460", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "6c2e2a3da84517d452ca79513f51a9cc6758619a"} {"id": "train_049461", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "edb4641cce223088a723d96b65c4a81fa37244d8"} {"id": "train_049462", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "673c3f86d6194efef79bd883c2370e13ece803e3"} {"id": "train_049463", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["security", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "1ddf7dc2543ab5763679e0ec35131c4c4ec6ba09"} {"id": "train_049464", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["security", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "18f138fded2d454b1f0adcd995076c640c4d91a0"} {"id": "train_049465", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "bfd9d51c689651a4d914a8fa6a224ea1e9be83ff"} {"id": "train_049466", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "a7b33cb192ea9b68902d3ea32657221551303ba1"} {"id": "train_049467", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["security", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "13dec38d420f7ddb7759dd437aeefde604f765fc"} {"id": "train_049468", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "b5aed132fc7a9dab81cdc3bd1b7d0ab1ecb71905"} {"id": "train_049469", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["research", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "710479b385359edd4a1140097b03cb115ba77551"} {"id": "train_049470", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "2e8031bfddeb26c6a5e94244db964094965fb837"} {"id": "train_049471", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["coding", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "6710e9f7ba3d23c5eb0f40c992ac9c59737914fe"} {"id": "train_049472", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["security", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "b0ff43a76af0ae35d583434edf65931423bcce4f"} {"id": "train_049473", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "783b3cb29716f883bb0282502c80cd8246dccc9a"} {"id": "train_049474", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "5a694566098a938644b6abb967d25fda71463248"} {"id": "train_049475", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["coding", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "8c44c0cd9f4a33529d66b73fd7768276602dbb69"} {"id": "train_049476", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["governance", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "7c8d265d7e7953f1fda8fbad7b21ff4966ae5f07"} {"id": "train_049477", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["evaluation", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "5a02f8e70cbb57a3dc37bdd435625f4440b42859"} {"id": "train_049478", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "1c17cd76eaed78ecf2c6a27e0de307e1827d85d3"} {"id": "train_049479", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Python", "moe_experts": ["research", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "ab5cf94eb459e80adcf7c4ad4a6916a57ea6d010"} {"id": "train_049480", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "14228746434456c3e54883cad34d17e834c54afc"} {"id": "train_049481", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n cost = 0\n for _ in range(budget):\n p = plan(issue)\n patch = edit(issue, p)\n ok, _ = test(patch)\n cost += 1\n if ok:\n return patch, cost\n raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["security", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "1db65ef5ead2b7eab92ea978afb73ca8349c2d44"} {"id": "train_049482", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "a39e80c91db68f9fe6cf87f2acb9d44e8d807906"} {"id": "train_049483", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["agent", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "ed1e491c4172f7fbe9ffae9ad43972ecd0d33ca9"} {"id": "train_049484", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["agent", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "f765b82e4bb37558b99063e91596a9914822f3a0"} {"id": "train_049485", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "4a7aebb47b959412f19429b5c302c688bd6069ec"} {"id": "train_049486", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "e7f58d5195f7efb8fe0e52d887d225fba2249f64"} {"id": "train_049487", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["security", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "263f74dbf129d00724ca89a9b40c13d6be10fd8a"} {"id": "train_049488", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["governance", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "6222e6fc2103e6803c3561b9da90ad04c0df2b37"} {"id": "train_049489", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["agent", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "366c667e83e68564812c987411a7ee8157d414fe"} {"id": "train_049490", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "93be4d3436d479c4cfa43ff40a0611051ecc2b36"} {"id": "train_049491", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n cost = 0\n for _ in range(budget):\n p = plan(issue)\n patch = edit(issue, p)\n ok, _ = test(patch)\n cost += 1\n if ok:\n return patch, cost\n raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["coding", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "ac56a1575228d2f6c82e1434e5a6fc94cf21b59f"} {"id": "train_049492", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["agent", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "efdd3a1640d3f42fe4609f761e04b054fe215b90"} {"id": "train_049493", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "d4140bb16522043d5e4df116ecd15a6ce7a361f2"} {"id": "train_049494", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["data", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "ee3c20846caa071cf9d8eda35749b39f2c2e21c3"} {"id": "train_049495", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["research", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "34fcc04a002e2ccd1e248b116ce06a2da7c34338"} {"id": "train_049496", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "6085c991db3f38b81ee7665570447a04a515fac9"} {"id": "train_049497", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "d9f25b1f1a862ea7b532a82f303f3dbe32326440"} {"id": "train_049498", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["governance", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "354fafb4890c3cb2a5a413fe5c9c9ecbe8b32be9"} {"id": "train_049499", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "8c14730b76d7608d83a188259a7ce1e90905dbdb"} {"id": "train_049500", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["agent", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "633033fdfb76d8e061c685dc92c9e067f9e9c13a"} {"id": "train_049501", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "f4cade7d8c8e935fa4cd3712ea8b5a6270c81b58"} {"id": "train_049502", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "2d609044a36b583893dcdb6ad607b2710d65824e"} {"id": "train_049503", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["performance", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "83e9e9fb018976ed41882aea07f1c40aefa2d862"} {"id": "train_049504", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["research", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "8f12157e17e3088c272bf8c213bde54ed3452bd1"} {"id": "train_049505", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "97a6ceb3908ba12ce9eb0d8495d31ab55142eb21"} {"id": "train_049506", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "5049860b1d7aa1795d0410315822b57adab9cab3"} {"id": "train_049507", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "8d128689c3a62e3de3e83bb7bfd223132d5e8987"} {"id": "train_049508", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "d49213776f545b45bb8e67aceaaf0d88a11fcd70"} {"id": "train_049509", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "2bb1898dbab77597103f2054181361a149dba176"} {"id": "train_049510", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["research", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "fc36369c85f3e8cffedb2cf2eaecd1f66eab546d"} {"id": "train_049511", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["governance", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "7034a84e72389ee14cbc62bb870daa92c4c5c4ea"} {"id": "train_049512", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["security", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "ad75e9133819f7ba7a6ad361e4f65b56973364d1"} {"id": "train_049513", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["security", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "e3d00bca8b0b29402d7808672c8cbfc9f50bd1a8"} {"id": "train_049514", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["governance", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "5538edea2d380aae9aa9917a50048d02b3e031fb"} {"id": "train_049515", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["research", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "5013d04ee9f95ab9be24b6b762d5e43ec777993d"} {"id": "train_049516", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["research", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "92bf970ea393b286ba8a3d9df93cc09bb4b6fc03"} {"id": "train_049517", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["security", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "877a36c8c2e02b5d9553a775879df51c3f5c79e2"} {"id": "train_049518", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["security", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "135c944f2cc4e9cb7ab938fb9153d299699aeda3"} {"id": "train_049519", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "C#", "moe_experts": ["governance", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "22e5322b20809d42fb090c25811b1416d0949d9e"} {"id": "train_049520", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Java", "moe_experts": ["evaluation", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "3475bf55a6badc4067cd4c2b5f92e2a7f6ac012e"} {"id": "train_049521", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "9a63a0ae61b9966f06740d23f08a8dcca06e1ba8"} {"id": "train_049522", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "6970355ba93f92c4928840b6fa9d5ce28cc0eb0d"} {"id": "train_049523", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["performance", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "38d52395a460afea6be2d2064b28791c68259aba"} {"id": "train_049524", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "433f61ccd1f90b9a01257c706d04ae99d7ac3a09"} {"id": "train_049525", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n cost = 0\n for _ in range(budget):\n p = plan(issue)\n patch = edit(issue, p)\n ok, _ = test(patch)\n cost += 1\n if ok:\n return patch, cost\n raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["data", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "4607d6d5357135e047fc12666effd1514887df97"} {"id": "train_049526", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["coding", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "cec11222e05d33baf1e563d218c601c3118572ba"} {"id": "train_049527", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "713060a4fff37b5f5996d4a371b6705d65634db0"} {"id": "train_049528", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "6c376405d100321d3904141d87d0b821e617ffad"} {"id": "train_049529", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n cost = 0\n for _ in range(budget):\n p = plan(issue)\n patch = edit(issue, p)\n ok, _ = test(patch)\n cost += 1\n if ok:\n return patch, cost\n raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["research", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "2284f6996618c39d73d52c604820e720c7fe25bd"} {"id": "train_049530", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["research", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "004b38693c97d4ce7477b0221c65fa25d7c37b1a"} {"id": "train_049531", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["data", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "14fbaecd2d21565b5551871163a646ea53ab2686"} {"id": "train_049532", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["evaluation", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "e343ca677344997a5554018c81cc6b280596fdd8"} {"id": "train_049533", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["research", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "25608dd20f959381f925b3f2d2f6e49306b9c546"} {"id": "train_049534", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "ac81d0ddda075c6662bad332dcfee5d3917229fe"} {"id": "train_049535", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "f68b3cff0afe292d404d38e4e0dcd2e35afe04e0"} {"id": "train_049536", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Python", "moe_experts": ["security", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "1405ac8baf6dcfb87b75620d0054a48febf33131"} {"id": "train_049537", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "cb6ff96f470954817af932b7b512c13e48740be0"} {"id": "train_049538", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["governance", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "790cc37eff4e1ba4c62137784683fdcc35a711f7"} {"id": "train_049539", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["data", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "a70aa8a69e0cb1acc055f9c5da32748bb81ae23f"} {"id": "train_049540", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["security", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "3c851c9b162d5b4a54ccb6374f48e354bf2bab1f"} {"id": "train_049541", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "86b38ef808c6bb1f43b82eefaae49366d7b0bc35"} {"id": "train_049542", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "abe643d049ca7227ace6d5c866b6dabba63f8e9b"} {"id": "train_049543", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "9b1ee3e4e2bc7d14450b5a16735dfe9846d1d696"} {"id": "train_049544", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["data", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "08ba4fb101460fd2c3ac39e3f921e9570833a0d7"} {"id": "train_049545", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["data", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "198b9678faa670b693d67e2cf0e052ecd07e3c55"} {"id": "train_049546", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "77aaf47b35fff9adf3321ed8625bb8b7c42428ab"} {"id": "train_049547", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["agent", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "1db73beeefa7999e70b6c8fa6d36a5208ce7769f"} {"id": "train_049548", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Java", "moe_experts": ["data", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "ccbb410d0735bcd82816866ce16b49274e2f1f29"} {"id": "train_049549", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["data", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "40e4b09c01cf34c2be4200c83c5c127c6f652f4f"} {"id": "train_049550", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["security", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "58fbde1fc0495fc3821077f04594741cdda4487f"} {"id": "train_049551", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "852aba3fc5ec5dc2406ada111a93248dc9ade8ea"} {"id": "train_049552", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["agent", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "6d2da016e62b792f9e808bdec69b021935228cd3"} {"id": "train_049553", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["evaluation", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "70c70c0cf50020032c0f683c8e9772d154337f76"} {"id": "train_049554", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["security", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "9d5f23987fc6fe524442fbcddfc2d4244e5c27a6"} {"id": "train_049555", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "1b1390ac273f03a252e8d2700a0cf2dab7c6671c"} {"id": "train_049556", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "93b7097e8776b9f3ec9fa920ef34681c45ad6e63"} {"id": "train_049557", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "400ed9801dd4fdf8bfb90082a6cf760027cb1421"} {"id": "train_049558", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "78239f9c1fc7839798e5fd2a814609e3061e38e1"} {"id": "train_049559", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "bf023b1946d406ae41874d2246534c92abba362f"} {"id": "train_049560", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "5736657c44ece8b63bc2db6e066efb9b0ada1870"} {"id": "train_049561", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "coding"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "893498c0a1623170d740b5a63c4726bdb47b22d4"} {"id": "train_049562", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "2a1828d9fa416f7ad9fba142adfe6f3d97e2993c"} {"id": "train_049563", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["research", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "feadd743ae6ab67f675d2a7de231d9ece2ca5af2"} {"id": "train_049564", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "7d9c30cac8f4adeda347ce5c63ec816290068438"} {"id": "train_049565", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "602b305216cb0024700a32c0ab1bcc85479de2aa"} {"id": "train_049566", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["evaluation", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "8d282358bd8ba590089c6eb663565597c36c69c8"} {"id": "train_049567", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["agent", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "bb2cfe7c5105e2cb6862b959cf4336a20c1cdce2"} {"id": "train_049568", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["security", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "e0dc4051717517018c1ed7d8d1570d5c79fb1cb8"} {"id": "train_049569", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["security", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "bb8067a72bbd7624f7333b40a588b69378abb41e"} {"id": "train_049570", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "af59f67bfb38ec0cc4ce73a47d51b33b2e56b90d"} {"id": "train_049571", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "92304c8619e170b4061e5fac0309e6251b3aac0f"} {"id": "train_049572", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["governance", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "4b75459e475b10d81c72a1ef73e06bfcbb42456d"} {"id": "train_049573", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "d721870d8b256604a55aea6250f4d82af59d9a72"} {"id": "train_049574", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["security", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "fb85c5617c14006522ef5a6e36ee6430d93e9f10"} {"id": "train_049575", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["evaluation", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "f018518258480b0c96e94387d5850785af2ab9f4"} {"id": "train_049576", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "41151031f3a7e6633d502a08214322bf64bda25c"} {"id": "train_049577", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["agent", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "471557ee00b972485fed12165d7e3009964c37ee"} {"id": "train_049578", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "5eb31728786dd7b789e59d8b6a943db3376a2717"} {"id": "train_049579", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Go", "moe_experts": ["governance", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "f867d366f34fb11b2df73fe7ff6a95b775347c4a"} {"id": "train_049580", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "15cae4c1cd5ca4015dabe7cb72189239b10c0197"} {"id": "train_049581", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["governance", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "aaef9b21fdd6991b091e2a6aee7fd307105da2cf"} {"id": "train_049582", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["research", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "4e96a277ec2d0f18eef305109e6ef5c0d749b83c"} {"id": "train_049583", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["security", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "51b378ea2e8a10002d175813779c6e8dce81ea78"} {"id": "train_049584", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "6d12a7e25c124cfef200af7e53966380100631b4"} {"id": "train_049585", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["security", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "3ca486007b978d0955ecda3197e72cc376dea17c"} {"id": "train_049586", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "50314f153af2fb2146e6775221f2c1e33871fbf6"} {"id": "train_049587", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["research", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "ab486dcf7f23ef7c8d77b37c4d0c3f10623724c1"} {"id": "train_049588", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "6ed221ff45ed373c151d0317df53cab02a53a134"} {"id": "train_049589", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["data", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "d7804ae9e6ddf61fbcce625cdca702196fc1e2a8"} {"id": "train_049590", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "103102daf600601567eab6ce31e97c97c23a3b31"} {"id": "train_049591", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "beb5acf4b3237394aa0e7491f423a7cb7fc89da9"} {"id": "train_049592", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "134939947da988224fb61bfe172145afdde90a2c"} {"id": "train_049593", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "b02888d2fb2af5bd5eac9502908d110cb2111efa"} {"id": "train_049594", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "b0647451eeb3b4dc85d02ad687a3544682fa5d66"} {"id": "train_049595", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["performance", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "708716251ae5898b7547ab9e54ad4ee18cd03cc2"} {"id": "train_049596", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "b9919990411185400bf5042a6967a0f60da589af"} {"id": "train_049597", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "2e0d68293e41311c0f054e0e23f625069cc880af"} {"id": "train_049598", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "996f29722e1bb35dad9be3fa9a99375d27cbbd70"} {"id": "train_049599", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Python", "moe_experts": ["evaluation", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "1bbeaf73337885b85907d62fef6d958fa11ed08a"} {"id": "train_049600", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["data", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "594309bf3f38a9e2d840d32da57b5a3163b4a808"} {"id": "train_049601", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "6c677604c39e47df42e2c72ec618e3e86fb3f413"} {"id": "train_049602", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "31ee8fed58856d237e597edc990f69ea0afd21ba"} {"id": "train_049603", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["coding", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "3f5fda589bf383a7942d0be4101bd72db4f84326"} {"id": "train_049604", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "0adbdf44c9d3b71c931bbe00fe36b2058b444101"} {"id": "train_049605", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["data", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "6feacc8e4c0b3af3becbe550f9afb1342c64edf2"} {"id": "train_049606", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["data", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "a944918a4e626fda7032218a01c0d87b0a722d5f"} {"id": "train_049607", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["agent", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "843c29af8b6cbf8947ba905ac0a8181ba69c9d36"} {"id": "train_049608", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["coding", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "ca4657b9ea66bda02f712529e5474bf83624f2a3"} {"id": "train_049609", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "027d82e9a3e537bf91b189700ce318cf9d7b206f"} {"id": "train_049610", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "1e9ef6a8ffcbfd89656461353886b14a316edf4a"} {"id": "train_049611", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "f4192401848a5dfb98e1826cd4efdfed4aa02280"} {"id": "train_049612", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["research", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "9ca76f6cc12a74a06c79689c7b5b3122ed252c62"} {"id": "train_049613", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["evaluation", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "88c50ca5eeb59494686a33d224392f0b427b8534"} {"id": "train_049614", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Rust", "moe_experts": ["coding", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "93b64cbd401bc01b76393243e0ed03a27a212852"} {"id": "train_049615", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "1257341cbe26cdf5a9bb74498477edadac1778e6"} {"id": "train_049616", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["agent", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "bce0bb1fa04719c743b967b6ab12b073d7fd5cba"} {"id": "train_049617", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["data", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "8c4c8e7000d98e38ae32883ab5b24c4fbc89f689"} {"id": "train_049618", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["performance", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "a92acad33605738574117103bffe5933e1ea7269"} {"id": "train_049619", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["evaluation", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "042b7feb19286302c0a2e7f056ec9b862c455843"} {"id": "train_049620", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Go", "moe_experts": ["security", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "6862d7f102f1225beaa089aca9087bbcbdb9756c"} {"id": "train_049621", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "7b4f4a94580028d2e2120365e449dbe66178a045"} {"id": "train_049622", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "5b505e2b78cca8fa4bced8c07fc87f8e78c5260e"} {"id": "train_049623", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "8d360de7f18d067ea53feb326a5a62a4988d5fc9"} {"id": "train_049624", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "689fb0b45e2dd13167b639de9967cdd5d543d77d"} {"id": "train_049625", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["data", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "175a7e3f0337c3aadafca5f26c0bc93e814b9226"} {"id": "train_049626", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "adea5592f1cb524d7d94f06063f00ad20b2b1afe"} {"id": "train_049627", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["data", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "ebd0a44b1fdbf5fbba07d9ad0957049228d4e942"} {"id": "train_049628", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "276e9431353bc86104d687df0286e0daedb5ca6d"} {"id": "train_049629", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["research", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "43b76f2f005177bbf499164d90a4b7a6dae53f61"} {"id": "train_049630", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "765529af724dad4a601efeeefc05ae364f9caa96"} {"id": "train_049631", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["research", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "8e103843d5380983a087c5dc76efcf7c71b7b748"} {"id": "train_049632", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["research", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "1faf6254e00a4eca538122ecbf563bac31b4d606"} {"id": "train_049633", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "0a6c412d51cf7e7f88cb68281cbdad5fd512d317"} {"id": "train_049634", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "9aa8be4d724af4edf7badd8112ec0843492293e2"} {"id": "train_049635", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["governance", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "c27634e9f36382d3e516056e01b41e5b3b08870f"} {"id": "train_049636", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["governance", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "bb0d9cedc3693ac4d974cab49a123c0b583517d1"} {"id": "train_049637", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "288dbba12cbed709dac13a719e31a3c65542f5ba"} {"id": "train_049638", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "329ad1c1390d3593a57d747fccf708b48fb751d8"} {"id": "train_049639", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "SQL", "moe_experts": ["performance", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "f42531544e87ddd9a00f9bf051c0697cd78294c0"} {"id": "train_049640", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "bec7d74ef22bd62983d9d8a6a3d59891dee36c94"} {"id": "train_049641", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["coding", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "4823aad2ecf4c2d30d88927bb6716105a90764c6"} {"id": "train_049642", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "2318c53f7c3821e39fd0c1f63c01463e91035d38"} {"id": "train_049643", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Python", "moe_experts": ["data", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "3981d817817db177007bdabce5827311fbaebf14"} {"id": "train_049644", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "d47514ebec3fbc5eb01b247bf3744b3e4244e621"} {"id": "train_049645", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["research", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "1a59ca3cbef91ef2fb8a3f49494e79ac5b24b06d"} {"id": "train_049646", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["research", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "76f76f2d04aae1f78aee391ff5cab8de08c4cf95"} {"id": "train_049647", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "8dd1dbd7e8f37e6674de015efb2e1eb718f10131"} {"id": "train_049648", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "491740f0f73f6519218ba8ec7035b9f7da723a5a"} {"id": "train_049649", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["coding", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "1f4efd3ab0ec4e255ad9011f2663c3f4067a516c"} {"id": "train_049650", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "8d9cf6a6a58df3ac5edec5659c1848290c8794c6"} {"id": "train_049651", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "9440882cc2046da254895f86201e4e92e53ae418"} {"id": "train_049652", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "evaluation"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "1c7f9c9242632d91c4ed83945798c26882d4658b"} {"id": "train_049653", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "3dbbd35bbe2cbf746bad3883b4ae9a68646dc0ee"} {"id": "train_049654", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["evaluation", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "681c6b7654018a4951d6e07e5128db0c41a95d83"} {"id": "train_049655", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["governance", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "f2bbded922cd9dbdf0e1c1b29ac61dd7f672cac3"} {"id": "train_049656", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "9f16e3c62c9660496ef62211879c36f105ca9cd8"} {"id": "train_049657", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["coding", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "42fea324ccea39febeee5f4ed9c510f134694e3c"} {"id": "train_049658", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["evaluation", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "dfaab52e21e66783106b486b099e432eb9142b1b"} {"id": "train_049659", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "100a5cede9532ed8925a2140ad7f6dc59f08003b"} {"id": "train_049660", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "84443dcc294ce31d9eed74f5b488a398bfe99f5e"} {"id": "train_049661", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["security", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "ee3c20846caa071cf9d8eda35749b39f2c2e21c3"} {"id": "train_049662", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["research", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "61654ca80ddc5ecd9a97e94518314270a5c31630"} {"id": "train_049663", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "5a4b3178382d07b16b5af6aa1633a4f44d56129f"} {"id": "train_049664", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Go", "moe_experts": ["agent", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "a8aa03c92fdbc3bd07db1f702be3dc761a82b8ba"} {"id": "train_049665", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["security", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "468de22d44bf8406aae41961a4385690b4e45aea"} {"id": "train_049666", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "3501f9ce735bf72aeb5904beae896a5f6fdd9abe"} {"id": "train_049667", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "ccd4f9b70fdfc256096cc48f37b3407c626b95d4"} {"id": "train_049668", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["governance", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "8b0ebfdfda7b469ffd42cc43b6600c188cf43df0"} {"id": "train_049669", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["research", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "67a5e99bb941f3f6ea1e7604a35a12094eea8c0a"} {"id": "train_049670", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "e9943bedea11c8d5922edff90d986ae3afede6eb"} {"id": "train_049671", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "321072ec108a0475034a422862d90f9afd4d5559"} {"id": "train_049672", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["security", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "c6af865555e57c7b12b7dfb12844787d0d8caa8f"} {"id": "train_049673", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "3c90900baa8aa9cdf5736020c2b35cff8c77e879"} {"id": "train_049674", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "75a32e7f3b8db0e43b81e22a8a2caa9cd4b99dd3"} {"id": "train_049675", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "7add556a8c542a476fbba624f38ef4dae5443746"} {"id": "train_049676", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "9ec0556e6dc16793a8b47e7bdea32dfe71810d38"} {"id": "train_049677", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["data", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "0a7eadbb3f2e6351aa2c71b9146cd85e950e3ee6"} {"id": "train_049678", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Rust", "moe_experts": ["agent", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "106219da6843a4610641fe7dfce7d62e341e9b13"} {"id": "train_049679", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Java", "moe_experts": ["governance", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "7b44ae94b7258459932cb83ccf16cc787f6f1f11"} {"id": "train_049680", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Java", "moe_experts": ["evaluation", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "7bad13b717b936616e9ee71ed2dae747294d1040"} {"id": "train_049681", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["security", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "c095fb0915413816949ef0b39b845b2788e185a6"} {"id": "train_049682", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["agent", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "bac31559ebd094d9799770a21eeaaa0dd93e7245"} {"id": "train_049683", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["security", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "d7f1d69f3ab2c01afcc8b61e34a30d6ec1953caa"} {"id": "train_049684", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "1c0be31b09a5d77e88f29d286ff7e900268ec7df"} {"id": "train_049685", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["agent", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "aa6c26611918924a00c7d427e89fc3c53d62806e"} {"id": "train_049686", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "65ddde9ac07963ca551be9a5c6b9ea906bcedf6e"} {"id": "train_049687", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["evaluation", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "0d4f6454c7750c5ed46223531b0fe04f2408165d"} {"id": "train_049688", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "d8b629654cd6c431eb0eeb3df1383d7b28abf73f"} {"id": "train_049689", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "be5923fbbe91f49592a1c265553b1296e46716a9"} {"id": "train_049690", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "fb640ea49a04c5347f656882df8d97e583614e8f"} {"id": "train_049691", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["security", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "8f233417dd449cae54aa135b9e10d38ed9ead24e"} {"id": "train_049692", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Java", "moe_experts": ["agent", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "20c132e3beb119b1d2ab68039f03c8d3cd97a862"} {"id": "train_049693", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["coding", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "51bf1cc2587254d7a8bd57596c60e291d256991f"} {"id": "train_049694", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "263f74dbf129d00724ca89a9b40c13d6be10fd8a"} {"id": "train_049695", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["governance", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "f785e2a1f58bff737d815a1380f492c651a6c1c3"} {"id": "train_049696", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "2bbfbc58324ac98dcd8112ebf33891fff63a2c6b"} {"id": "train_049697", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["agent", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "3269c41205f4bd2997e19e5e3d59f507189c83cc"} {"id": "train_049698", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["agent", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "0aa1e3c7d0374b8e8cdf44763dc4dc72a38b4907"} {"id": "train_049699", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "e4bdf19152541a61346c1354bc6c157d8d98f579"} {"id": "train_049700", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "87785d212ff2d46fee7308e60f09ff2e2802a25f"} {"id": "train_049701", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["evaluation", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "674f2f6cb1d8cc7f77c67a1941f9929e05b66f46"} {"id": "train_049702", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "2b59c3068a99fdfa99bd6a0b96e4da2e651990a7"} {"id": "train_049703", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "0039a6f98bb725bbc0f0c073e8ca8c742f3f02b6"} {"id": "train_049704", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "1073c3e2a533755b9f7dc5c7eea93a622cea586b"} {"id": "train_049705", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["security", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "1591c8d42b4bc99b80252843d79418e6596a5972"} {"id": "train_049706", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["research", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "e7440794621bd5acb5172631700b2129f89546c6"} {"id": "train_049707", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "6ac07b978286d6e8b88ca0090f34e076323ce979"} {"id": "train_049708", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["performance", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "1d45f9a6e8787c5d885abd7d3a970efc14965804"} {"id": "train_049709", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "da8bcc1883d19a730360ec3df08e3c71b2cff5c0"} {"id": "train_049710", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "5b6e928dd39ad43643a19aa7ce79acb995dd78e5"} {"id": "train_049711", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "d167ef22f2f65ec5b387700cc64276d1a266542e"} {"id": "train_049712", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "e2b67df2b04b911826517f0a6b6f1e2049284516"} {"id": "train_049713", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["coding", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "584ff6f5fda023eac22c8b73d4ff5eac3ab1d9ef"} {"id": "train_049714", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "e209c4d459d2356a31f64ac3786bb70313daad6f"} {"id": "train_049715", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["coding", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "5c6b21a7177794f3c1b9c23d99dfbc15f9725d96"} {"id": "train_049716", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["performance", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "5218befb4f3a9513b9013455bcc919d82fda8e84"} {"id": "train_049717", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "a0bba63280841063c47f433dff0f42271ee9beb6"} {"id": "train_049718", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "5a781bf9ef2b442071ba2fe89f41d460c6f1a007"} {"id": "train_049719", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "afa6e9a5970d52d1b1b636ef9f6af777dee133f4"} {"id": "train_049720", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "cbff56cf82b7fde9c92ba331d0000fe8126f9e82"} {"id": "train_049721", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "a730905bc81e54ece1c8902705f5e1d9fbd21f4b"} {"id": "train_049722", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["governance", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "beaea352df8b64f719be3d7c2e3286f17f19678c"} {"id": "train_049723", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "5a88e5e67e2974523638146b7089cef56083a1be"} {"id": "train_049724", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["governance", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "85af49378f23a15ed73038618c181f7f9902c44f"} {"id": "train_049725", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Bash", "moe_experts": ["security", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "6a0f4b8a44c8aa0ae3b725a6543f6c09250e528a"} {"id": "train_049726", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["evaluation", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "d128a99853eb78d08eb6621ac187811480dd7dd0"} {"id": "train_049727", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["data", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "cf44506cf60f5bef7d20f8c5321b6ca284b7553a"} {"id": "train_049728", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["agent", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "95903ba63c033720893bface06e1edf4905a0a68"} {"id": "train_049729", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["data", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "321e26ed591da28c2da72252feba4813d41ffe38"} {"id": "train_049730", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["data", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "57655d196b85ab47c0827617bdf66ed191ecd900"} {"id": "train_049731", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "8a478fefdd52e31e56fdab31960d882c53938ee3"} {"id": "train_049732", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["data", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "5ee43c6c3d8e11a87163183772b67792ba8df8a4"} {"id": "train_049733", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 9,\n \"confidence\": 0.9,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["data", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "a2c7bc2ca063ed397d076fc91d0258f8d6d7b444"} {"id": "train_049734", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "22f24e78aa1f427f3478c77a459ed525f60a95eb"} {"id": "train_049735", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["data", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "9ea42c3b29cb1bd3a7c17ac3df99b32e51c2014f"} {"id": "train_049736", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "75ce1f13dacf3d1f7314cbbe4f2b588e0346af66"} {"id": "train_049737", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["security", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "405f02a55fafee62808440eceb280add7b51068c"} {"id": "train_049738", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["evaluation", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "14f4f6b84fdeb73222ad7bae75a464d78135040f"} {"id": "train_049739", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["research", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "f97ea717caa30943a7bded196b55b56f9336c82c"} {"id": "train_049740", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "bd52cca696f51c96a25df6a13c2f4a79672a9f9c"} {"id": "train_049741", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["security", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "da2b0d59545bfb558d47da91212753fb6589a17e"} {"id": "train_049742", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "8f7f19c7c1d939aab0683addc408ed3d2ee480f5"} {"id": "train_049743", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "f20a782714259dd22f754846cd337d5055b86d19"} {"id": "train_049744", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Python\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["governance", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "7e50921d8706b5b53c6b3638a5b1523a90b1b7bd"} {"id": "train_049745", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["research", "security"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "364005fa62dc2035cc8d62a783aa96280a1ef3d1"} {"id": "train_049746", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["coding", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "945c3a3c3a6c28cb1ed367dce93901022bebff59"} {"id": "train_049747", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Java", "moe_experts": ["performance", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "b9c269d1397fd27fb259c7414b6cbf1a1bca4bcf"} {"id": "train_049748", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["security", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "9fd7cee51ae7564300676403990b13dfff60fbe6"} {"id": "train_049749", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["evaluation", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "f823ff5166744adc330681c19909810581d16437"} {"id": "train_049750", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["evaluation", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "bb8067a72bbd7624f7333b40a588b69378abb41e"} {"id": "train_049751", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["security", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "af5ed1cc5b98123f89087db04c7c92861a15f843"} {"id": "train_049752", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "agent"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "50f6b4300d1d598922ab97043f14de16a0ae75d9"} {"id": "train_049753", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "SQL", "moe_experts": ["data", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "44d0cd699276beaafa9280b7a671b584325b9363"} {"id": "train_049754", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "evaluation"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "ac796a832868d00b8f93a39cf7cc4a44dd6e8cb8"} {"id": "train_049755", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "8116c679af5fd0d79bb2158b47fad006ad826e0f"} {"id": "train_049756", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "68f267ab99b3eeb486ebc974e1431c3c222a99e8"} {"id": "train_049757", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "security"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "b5742767a969b268dc89d93781c31aac5b54f5cc"} {"id": "train_049758", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "d296e7e77079873d186204de0af718ce32131b70"} {"id": "train_049759", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["governance", "performance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "ff959f543a02d8bc15fce72e41f9e9459edd392e"} {"id": "train_049760", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Bash", "moe_experts": ["data", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "dd9546b0e846b22777430cc559372afba2372e66"} {"id": "train_049761", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["agent", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "1c17b349c863db8d5518d0328310ecb9fe9654d6"} {"id": "train_049762", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 9,\n \"confidence\": 0.9,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["research", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "ab795c2ef1adf4f871f5d601c8d711922eeed18b"} {"id": "train_049763", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "39707bc5e84a98352b84657ff08e85560790141d"} {"id": "train_049764", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["coding", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "048bdb5093ec68a0a72eda0ad63a92b3aaf9077e"} {"id": "train_049765", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "aafcd12689ff6f0b887bf3970f24e4be7007488c"} {"id": "train_049766", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["evaluation", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "1ae7ad7f12ffe54f0f5b7359726f952b4435a9a3"} {"id": "train_049767", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "5909dd5e88f99a6a51c52a4bbf145dbac3643863"} {"id": "train_049768", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "f9d2a9103ac9b8fa29ece6659fe001f3728c8748"} {"id": "train_049769", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Rust", "moe_experts": ["coding", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "40f7228171eb11d09c8449f70db4029bee6895f9"} {"id": "train_049770", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "418ed4a4c9d540c045010bec02f4ee2341f3ff36"} {"id": "train_049771", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "a83fa917a3a09f1614488ee59e7e591ab13ca754"} {"id": "train_049772", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "8621a6709b5a5ac5f48d1d12fcb3c46645e0f35b"} {"id": "train_049773", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["research", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "e2430f05c3b2712158dda100db24743c0b72ece2"} {"id": "train_049774", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["data", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "53729681b1668f729ba136dcbfa57b9c3fb0c1b4"} {"id": "train_049775", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["governance", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "633b36748e2b7969ab3bc3d5e02bf63ce3294357"} {"id": "train_049776", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["data", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "2de24f3018c2218734bca261c99a7f5bbc3af0e3"} {"id": "train_049777", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "1ba1cb6a0a0ab341833c2f503716c42ae9e85c35"} {"id": "train_049778", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["governance", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "7f2205fec21a3c627219b6a5e7c009141905f719"} {"id": "train_049779", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["performance", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "ccb9cd391fb3a4737412fefb49803069ce081894"} {"id": "train_049780", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "b9283878999ec42dcd8c864e4b847ec36942cff6"} {"id": "train_049781", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "69ce7bda7e1f1afca9b8010192723205e0397cbc"} {"id": "train_049782", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["research", "evaluation"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "2498980ab6767204e168e526b681606149cc3d10"} {"id": "train_049783", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "b64bfd1d4100975aeb9babd500f5286c94fd0a0e"} {"id": "train_049784", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["security", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "1cd56d5d8155f9af63f34d441a04d3743d874614"} {"id": "train_049785", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["agent", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "ece02fea245f32e47ce3a4019a572ae9f82ccbd9"} {"id": "train_049786", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["governance", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "a1283754789ddd5ebb8c31c1323f557515de2af0"} {"id": "train_049787", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n cost = 0\n for _ in range(budget):\n p = plan(issue)\n patch = edit(issue, p)\n ok, _ = test(patch)\n cost += 1\n if ok:\n return patch, cost\n raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["security", "coding"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "91fbb42b2091872eb47ef6665dc201c4a5b8e801"} {"id": "train_049788", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["coding", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "187c367208e489acffbc283f13ba94b6b050efad"} {"id": "train_049789", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "da835465794a8cccf6efbeb0b24b2d62da833601"} {"id": "train_049790", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "83a68b46bc5d72038d27016d35361e78636764c0"} {"id": "train_049791", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["agent", "security"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "c8e8462bbbc5ca142e6ed6a99be285a59ab3b0bb"} {"id": "train_049792", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "3dc53d1be8229ab647553adb34d1fd2ddbc1b9de"} {"id": "train_049793", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["data", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "283fc177fc5a7e58c9de3f9049b47f959ca8163c"} {"id": "train_049794", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["data", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "3a12ac0e36a45b7e5a30b734d5f910442914429c"} {"id": "train_049795", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["security", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "cebbc4b537821083b11bf0a6191e5debfcfcf433"} {"id": "train_049796", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["agent", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "7335312f5104e0945d5afd5e1d199172f9bd2cbf"} {"id": "train_049797", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["agent", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "257eec780851a5c1e33ca31d012413bb7974f7ec"} {"id": "train_049798", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["research", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "48d60c823df17306c1ccebe12c38b6c1bc03e1ad"} {"id": "train_049799", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: JavaScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "07553415fa80a7a82e1053504056d6eb1bea7d05"} {"id": "train_049800", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["security", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "e297906f2977f7427d105ae59a12a85552b0aacf"} {"id": "train_049801", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["performance", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "c54d7100666aead0ed84223da9bd0d548bd4e68d"} {"id": "train_049802", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "SQL", "moe_experts": ["security", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "0667dd2a816890e0994eadb7c5346534f14b70e0"} {"id": "train_049803", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "8426b927961f98bdf9fd64c85e2197550d6d04be"} {"id": "train_049804", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "performance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "d974f4ef95f61ad0b2ba2a5bdb0ca9acf4858bc1"} {"id": "train_049805", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Bash", "moe_experts": ["performance", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "566e357c47b861a4ec027fc05152c5c703b5185b"} {"id": "train_049806", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Go", "moe_experts": ["agent", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "e156fac250e4e91d405161d4d943c3303b7dd1f8"} {"id": "train_049807", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["governance", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "014573266554a3ecc3d4e0f42d5b691788838e1e"} {"id": "train_049808", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "SQL", "moe_experts": ["governance", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "40714721e02432ab6a7e601ad6e9641528a2b27e"} {"id": "train_049809", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["agent", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "9e21d90da0959fd5050eccfa4940ef6323a0c5bd"} {"id": "train_049810", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "cc6036d68f774d2b3fcbfc28bbe5c2821b9c90c4"} {"id": "train_049811", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "95436c3a24f8b2e82249525082bb88297cd67ee4"} {"id": "train_049812", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["evaluation", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "3eff1572354ea78aa37096a6825d7c488eedba0f"} {"id": "train_049813", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "SQL", "moe_experts": ["data", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "e13b3ea6fd8ea44faa65216f4e0d3d1c30d3801a"} {"id": "train_049814", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "c6953f7c6e2d7532e2458cfb457913ebf6334500"} {"id": "train_049815", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "SQL", "moe_experts": ["coding", "data"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "dc97072dcfd61bcc2bfa2e059b5a42ff75f4486c"} {"id": "train_049816", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "861b73950b0fc596e25e0e6d34fef07e4a7894a1"} {"id": "train_049817", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "performance"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "a38b7aeace3070e0797f5c27ddfcc67be1cfc0f6"} {"id": "train_049818", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "177448a18483a1b93b937dbb703ee22ac9a8362b"} {"id": "train_049819", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "bd12f9b46489ac8edfa806efc17bac0d1e1e601b"} {"id": "train_049820", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "45ada70b3538f0385104bcbda4b3f3ca8fefc352"} {"id": "train_049821", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "be8d9efbd70b16fe834e925ace6f0922c70a3620"} {"id": "train_049822", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: Bash\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["research", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "c54c7a17d9cb78635af00fcde964695f94fcb5a3"} {"id": "train_049823", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "9916cbd38ab1ebe6e30e69f43370c750ede438a7"} {"id": "train_049824", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Rust", "moe_experts": ["coding", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "d43db0585ca2f41fe3483b1036d0fc5e2d4c8d1b"} {"id": "train_049825", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "SQL", "moe_experts": ["data", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "472442fc9ac65abd7f87c17814d75fec0c3e417f"} {"id": "train_049826", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "4a5c4d88cd0130607186ce5cc7f2cc3d8f784eb2"} {"id": "train_049827", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Python", "moe_experts": ["governance", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "33a032b1c2abf5c6527266460cb36ed66b3fbaaf"} {"id": "train_049828", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["agent", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "7d6f926dee313c357a51e8ecc5e532a91d49c8c8"} {"id": "train_049829", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["coding", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "6b439307e49113b836e72d1429f50837336d0d76"} {"id": "train_049830", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["security", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "afdac676b1d779dccc8603ba9450ce7f245308a9"} {"id": "train_049831", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "74dfab6e55e80c24fa784de70a256617452535ca"} {"id": "train_049832", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["security", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "81a48c1bf2d9c068f47104dd98f58f9f0730e1c5"} {"id": "train_049833", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 9,\n \"confidence\": 0.9,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "497cb74263202b65e76b663b703e1d59c3168be5"} {"id": "train_049834", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["data", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "aed0d78aed901ee9f567dea09250d354a3a46f26"} {"id": "train_049835", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["evaluation", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "b58a2cb496e0b4aaf1072a76baf60c3dae81c2fe"} {"id": "train_049836", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["performance", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "7ff2fbd7c9ddc42c68908a23d53d803be8ea3979"} {"id": "train_049837", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "7383d84da02973f856f419e7273e5bf42d1295c6"} {"id": "train_049838", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["governance", "evaluation"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "3da06adc8ca459ba89dd01724aa858533e244661"} {"id": "train_049839", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["performance", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "ef9450f1fcb77010e96aba3daa09c598f105b709"} {"id": "train_049840", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Go\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "agent"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "a8e9f5ee8f57e09b6cd3b3135d6e0c0d41b895fd"} {"id": "train_049841", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["research", "data"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "9b6077b0d6598886e25650b81986d4b437cfc5aa"} {"id": "train_049842", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "agent"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "234058a11fd0fdb1a307bc8c0f3995c8fa775b91"} {"id": "train_049843", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "governance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "257eec780851a5c1e33ca31d012413bb7974f7ec"} {"id": "train_049844", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["agent", "coding"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "3c3b89e9be57a31d42372f0a9996afe792bd765d"} {"id": "train_049845", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["performance", "evaluation"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "b4f8ae07aaac32f086c4256be8d78337e96748fc"} {"id": "train_049846", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Java", "moe_experts": ["security", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "a84b4694328591abf9342a67881e7a7bac73f686"} {"id": "train_049847", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["agent", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "c72995f8074bded8348c8714769095cae1af3638"} {"id": "train_049848", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["data", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "90ab42dad605a83211eea027cfeb9c45966ec948"} {"id": "train_049849", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["agent", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "b8b910aca9a7081194588b08c81869f4b092cf02"} {"id": "train_049850", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "c4355b5e6400a05eea1c1bae977b251ceaf7a2a0"} {"id": "train_049851", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "SQL", "moe_experts": ["security", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "46a6d0f71216c2e7e22301bee3ae81de9706c8b7"} {"id": "train_049852", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["coding", "security"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "c1e823576ef56dd1027c07e8bcd2cf31dc6adf71"} {"id": "train_049853", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["data", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "26af262c46fe82c80a7e85a125d68726d3cdd427"} {"id": "train_049854", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["research", "agent"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "be6382eadea95bab6807f4b0c6e833afe4ead607"} {"id": "train_049855", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["performance", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "be6dce8fa9d465c97c4a7bb76255c00868bbcb6d"} {"id": "train_049856", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["agent", "performance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "4de64016b693b04816eaf1aed2c86e64aab56b74"} {"id": "train_049857", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["agent", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "55929b9bfe723bef1e2760eeb5f4ae8904d6cf26"} {"id": "train_049858", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "f028cdc59cfb28c93fb6af194b954a91e7828fa2"} {"id": "train_049859", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "f41b28dcb95b3e4cc45e21bf9f902b41119582c3"} {"id": "train_049860", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "C#", "moe_experts": ["security", "agent"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "32ab6fa094f3c8e97417cddaa1401469149c270f"} {"id": "train_049861", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["coding", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "af3978b6506f351db3d3bfd38d3c0ef4cc2c410c"} {"id": "train_049862", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nCode:\n```python\ndef agent_loop(plan, edit, test, issue, budget=4):\n cost = 0\n for _ in range(budget):\n p = plan(issue)\n patch = edit(issue, p)\n ok, _ = test(patch)\n cost += 1\n if ok:\n return patch, cost\n raise RuntimeError(\"Budget exhausted\")\n\n```", "metadata": {"language": "Python", "moe_experts": ["coding", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "381d22234ae948a98c3a6aef0c3a18e0cb1c325b"} {"id": "train_049863", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Failure engineering and resilience testing\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "evaluation"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "4857dce2c7800141f56d329218bf4fd2ec0787dc"} {"id": "train_049864", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["governance", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "f0e4d2e32be86a4005d8aa3aba3b183145034df2"} {"id": "train_049865", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["performance", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "c3c326a91418d3ff95bc91caea0bb5430e8bfc1d"} {"id": "train_049866", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["coding", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "eb5635cd7b9886732a510051ef4a39f47b98dec2"} {"id": "train_049867", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "78e15df2b90989489cbee52809e8a2cdfe7fa916"} {"id": "train_049868", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["data", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "e9a26e37bbc6cf5beb7e0d7f81264f32e5361e46"} {"id": "train_049869", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "45b79ac09049c1ecebf85437dbc4a3b3b8087627"} {"id": "train_049870", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["coding", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "f071ff87e22659f730329f0f93541a25f40a5a50"} {"id": "train_049871", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Bash", "moe_experts": ["coding", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "90e64ebec8e27849dc0876854f2d87827899bd88"} {"id": "train_049872", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["agent", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "50e632ccb113e8b114bee92c64987f9420dc58b0"} {"id": "train_049873", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: C#\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "48c1d814c98769ee8bcb25ff6db34db71bd66437"} {"id": "train_049874", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["governance", "research"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "24d574dd29e5743d0b9f7d6b058fd27b4d98b1c0"} {"id": "train_049875", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "C#", "moe_experts": ["agent", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "88b84cae9765369fc51b273aec0a29cdd61fdadc"} {"id": "train_049876", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["governance", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "baa9af1e9148a9d2d8805b3d7e7b3f503eb17e84"} {"id": "train_049877", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["security", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "4750a66eb4ac4b0c71b9870a2fee97f5ed4a87f9"} {"id": "train_049878", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "coding"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "c265efdba89955d2f6fc1625b101f89e0d6d6874"} {"id": "train_049879", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["research", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "17f6f2235e8ca97ecd55aac6321075fef4117954"} {"id": "train_049880", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "10d5336b3dd1d004b5fb63177ec1e6567757a0c1"} {"id": "train_049881", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "f2383bf94c102a44d155b6214da74278b3e88824"} {"id": "train_049882", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "research"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "f14b9873e30ce7fa28e21681255e9bc59b01a1bb"} {"id": "train_049883", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["coding", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "775995559d54a596609aa0440fafac171ad4fce5"} {"id": "train_049884", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["governance", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "6f06c5e28e27978c520f63b9c1fc2e6c57b7a353"} {"id": "train_049885", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Extended context, memory, and repo graphs\nDifficulty: expert\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["security", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "ee3b1909ef5f07a8daff379ecd490dfb33fce1c8"} {"id": "train_049886", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Code-specialized model families and scaling laws\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Java", "moe_experts": ["evaluation", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "506705fd42551ca4c8e0902516d2f2ec07680fa6"} {"id": "train_049887", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 9,\n \"confidence\": 0.9,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["research", "governance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "6e5896dfa0d798f552690116a1ffefd31d1c8258"} {"id": "train_049888", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["data", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "8a879ff6615fb0c846a680ce8c73bb40f5372ef0"} {"id": "train_049889", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["agent", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "56868fb8e7a21d2c4e7d33900e93a5cce86ec8b2"} {"id": "train_049890", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["research", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "7f7194a6f43fb36c7883cf133fac4168b8658fb9"} {"id": "train_049891", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["agent", "security"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "3e595c05440d7a300c9acfc5a75a42ed5788bc05"} {"id": "train_049892", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["agent", "evaluation"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "28438f3ce22d0a4cd5c45eb944bc66a63558498a"} {"id": "train_049893", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["security", "agent"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "ee7e4de174ac5676892f115e0f15e1c761c2b356"} {"id": "train_049894", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["security", "governance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "90bde15a9756cee2ee4a604587a2b129de212b67"} {"id": "train_049895", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "TypeScript", "moe_experts": ["security", "data"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "65d996ce87d918309595e6aeca3ce88a3750ca98"} {"id": "train_049896", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Go\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "f76bc4a772d07d21b5e821457936f0b12a91df29"} {"id": "train_049897", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: JavaScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["data", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "1806ebe8f8da2aade4a9780c111d4f81f6becaf2"} {"id": "train_049898", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["coding", "performance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "cc7e5d5f30b32bafec946cf0c6057b007a844519"} {"id": "train_049899", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["evaluation", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "b5177fba56ed3e48757730f4c86b6f1e2d461c4f"} {"id": "train_049900", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "performance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "9441d2a9a1e9fd02b65386ccf065cddee7e58f57"} {"id": "train_049901", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "c83ea10916edd6952610ef405d2afe1082f5767b"} {"id": "train_049902", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["data", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "8f1e62aff4aa092abb7110df3896e0a52bd6306c"} {"id": "train_049903", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["coding", "data"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "dfaab52e21e66783106b486b099e432eb9142b1b"} {"id": "train_049904", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Java\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["research", "coding"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "fe1a4fe8a4d2bd43d870c90e4196723b3f19472c"} {"id": "train_049905", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "C#", "moe_experts": ["governance", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "0f7d118edb8aade090f327b2d007fcbcf3bd4f6e"} {"id": "train_049906", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 9,\n \"confidence\": 0.9,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "JavaScript", "moe_experts": ["research", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "a77eec75e3e385301796dc2a75e143b84edc64c5"} {"id": "train_049907", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "performance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "57a83e55a6a221e378d88b2263b679ae20930ef4"} {"id": "train_049908", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "ac6a0a5ab593a4f24d8c1af61842547815a3992a"} {"id": "train_049909", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["data", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "e4a2143c0bffee7aaa147bc4245b9a2027b0155b"} {"id": "train_049910", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Rust\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "coding"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "278da8f2fa3ff2f998b452c849dc5267a7b50f89"} {"id": "train_049911", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Bash", "moe_experts": ["security", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "bc511480d625049aa6b3978cdf6fb725b6cd1a17"} {"id": "train_049912", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["agent", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "cbfe290204d2a70997084b5ebf0c3b561393192d"} {"id": "train_049913", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "7cc21443b619546dc0fb22f81233ae3e0518b808"} {"id": "train_049914", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["performance", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "231344d7d4e3a0c7aed09b436137715b135bcf83"} {"id": "train_049915", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["research", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "ee9ae1c88a266d34f8be11002062c3be897573bc"} {"id": "train_049916", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["coding", "security"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "54c3491b55c435b0b4be9edeb21c85e7ba7d04ac"} {"id": "train_049917", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Go", "moe_experts": ["evaluation", "coding"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "508822e4e3ca299c43e49f44bcff3417e3826ad9"} {"id": "train_049918", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Python", "moe_experts": ["security", "data"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "1b1dc26e6603f665e7b436a8bfdc11b2a00542d3"} {"id": "train_049919", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["data", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "3c58f13f57b7646cd518faaff34f24beffa5fec3"} {"id": "train_049920", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["security", "research"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "1261b76225489149508ba0be92aaca79bb36c0dc"} {"id": "train_049921", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["agent", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "361fe4ab921dceeb91530c44a2be934154114336"} {"id": "train_049922", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["performance", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "af59f67bfb38ec0cc4ce73a47d51b33b2e56b90d"} {"id": "train_049923", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["coding", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "4af3701aa3a208bd5e7349876ec93ae203d32c6f"} {"id": "train_049924", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["governance", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "71b1a955b1d9c74128c4e8575e4aad7c154575b2"} {"id": "train_049925", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "performance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "56fcf6dfd06ddfcffadd2a2f775bf41cd955161a"} {"id": "train_049926", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Multi-agent coordination and role separation\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 9,\n \"confidence\": 0.9,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["coding", "security"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "14c3790b92d78d621aaad92496a36ee34993a814"} {"id": "train_049927", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["security", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "bc6a7f06eb30155ba39fe2b8af86d4ff8ecb24dd"} {"id": "train_049928", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["data", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "4f7e800b6b272fbb4c7a34c0a4ff9916e01b33eb"} {"id": "train_049929", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Python\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["agent", "evaluation"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "4c50abb83001962463f1692cbfb6df80eed90a1f"} {"id": "train_049930", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Model merging, distillation, and continual learning\nDifficulty: advanced\nLanguage: SQL\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["coding", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "0b7bbd1869c9c76772a365dde62224ebc3030e16"} {"id": "train_049931", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "JavaScript", "moe_experts": ["security", "evaluation"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "29db0db86b7e26d6aea894f70cc5346a7534f5b9"} {"id": "train_049932", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["security", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "1768f522d457327c9e03c9f23c016234e5ab1640"} {"id": "train_049933", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Self-grading and automated model critique\nDifficulty: expert\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Rust", "moe_experts": ["performance", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "847e4030a0d3b7a8cef1bf2034d71ef04671bd28"} {"id": "train_049934", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "a2f687c391967a7cb763c965c12425b730e7b1d5"} {"id": "train_049935", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: JavaScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["research", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "ee19fc94891da6c74426fe9a626ba7a7b0681c33"} {"id": "train_049936", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["evaluation", "governance"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "7e26da5a7edfe8ace3bb98a5cb922bcb44ea6ab4"} {"id": "train_049937", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 10,\n \"confidence\": 1.0,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Java", "moe_experts": ["performance", "research"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "ed52fbc80eac52d25be9ce2d43b60b987ed51ef6"} {"id": "train_049938", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Java\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["performance", "security"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "c716715859fa4f18eac6dc579f6faf4a6baf735a"} {"id": "train_049939", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: C#\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["coding", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "483aa43edad00b8df0814b6964a521184b87b695"} {"id": "train_049940", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Go\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "coding"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "678acb45823105ce0bb8e85416aef939cc71d488"} {"id": "train_049941", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["data", "evaluation"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "54e51bdc5d8e9b6f2a40f68ea5d562935ffcf61f"} {"id": "train_049942", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["governance", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "6b9f5f817311b00d72215e2c6dcbc81d57cd8d2e"} {"id": "train_049943", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: C#\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "research"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "492faa2de733c1b3f675477adcc1e6e526f3a4b6"} {"id": "train_049944", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Python", "moe_experts": ["research", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "4baade02d402298f31ce0f6ee7881344a2338639"} {"id": "train_049945", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: TypeScript\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["security", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "73b410e3438b369ba63fd4071cf73adf6579410f"} {"id": "train_049946", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: advanced\nLanguage: Rust\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "321b2e5cdd2cfd47828bd70c1684f342871b049f"} {"id": "train_049947", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "research"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "3dc0cdbe9491b6d2d0bc727b0d98ffe57fae6cfb"} {"id": "train_049948", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "performance"], "cost_budget": 3, "latency_ms_target": 50, "audit_required": true}, "hash": "1c604b0e428fe2e2d43155ab4a65bc62773b935a"} {"id": "train_049949", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Rust", "moe_experts": ["coding", "governance"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "3841b94b3575664e1ce5dd04e88c267eac01e250"} {"id": "train_049950", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["evaluation", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "87aafed8f3d8adf2408f9057ff4ca39aa1a947ca"} {"id": "train_049951", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["agent", "coding"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "020eb2282a54f79440757e44fc933556ebf9eca9"} {"id": "train_049952", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "security"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "f93eaa471a7e4b0692b1830c206192b339ea08eb"} {"id": "train_049953", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["data", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "f67313aef7d53b24ab20cdc15b38586ba0659ce1"} {"id": "train_049954", "created": "2026-01-01", "topic": "Code-specialized model families and scaling laws", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Code-specialized model families and scaling laws\nDifficulty: expert\nLanguage: SQL\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "SQL", "moe_experts": ["agent", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "52ec9cfc5507f38985e9d583a402a0cd7394b9a0"} {"id": "train_049955", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["governance", "agent"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "53e97288630ba39f8c46cdec4a7ce52e560071cd"} {"id": "train_049956", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "agent"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "1d4061b0afb028f3b9b12f45113691dbfca92765"} {"id": "train_049957", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["coding", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "6d4f15dbfa0c9d39b8c77838387813c5573ebf0b"} {"id": "train_049958", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: Bash\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Bash", "moe_experts": ["coding", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "5cb558a0a49d6a6fa3d77c9a52739bac44e76c2d"} {"id": "train_049959", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: C#\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["security", "data"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "f29c2d364cbf84a2b9b96d87f340c6b5f2d9c3ac"} {"id": "train_049960", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "JavaScript", "moe_experts": ["governance", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "a75dcd5e593d028e53685ac02895073573de5bf4"} {"id": "train_049961", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "performance"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "0a58a9dd92c407cb9aeded875896969443396f7a"} {"id": "train_049962", "created": "2026-01-01", "topic": "Multi-agent coordination and role separation", "task_type": "tool_trace", "difficulty": "expert", "instruction": "Task: tool_trace\nTopic: Multi-agent coordination and role separation\nDifficulty: expert\nLanguage: Java\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "Java", "moe_experts": ["agent", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "6d98e5d299d69e639e9ca9ba0b067597be3c4037"} {"id": "train_049963", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: expert\nLanguage: Python\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "Python", "moe_experts": ["evaluation", "coding"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "282df6268bc57e4f706734fc18f82fa0e6dcd667"} {"id": "train_049964", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: JavaScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "JavaScript", "moe_experts": ["data", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "738f2ad90c5ccdef9947014b349b23aba521d52f"} {"id": "train_049965", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 6,\n \"confidence\": 0.6,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Go", "moe_experts": ["data", "research"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "25edd9ffcb95e249a2ba08812bfff4dd9dcbabb9"} {"id": "train_049966", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Latency, cost, throughput optimization\nDifficulty: expert\nLanguage: Rust\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["research", "governance"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "7f0cb32cfb3cb116f0cb6eb2dc4218a40ce97f49"} {"id": "train_049967", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["performance", "security"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "4405366d1e1e1eb2f19385f7e9d19720062fdefc"} {"id": "train_049968", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Python\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Python", "moe_experts": ["agent", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "92737220867ad8f92d0e3ddf00027efeeb35e3c1"} {"id": "train_049969", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "TypeScript", "moe_experts": ["security", "research"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "123798e5b579189329316b8067ae73adaa26287b"} {"id": "train_049970", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: SQL\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "SQL", "moe_experts": ["coding", "agent"], "cost_budget": 3, "latency_ms_target": 200, "audit_required": true}, "hash": "560259e481d4a94e23f87f2c69f10de42e35c3a2"} {"id": "train_049971", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["agent", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "a9756f6f9a48cc18206a2aee27f6f998ffd3b35f"} {"id": "train_049972", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Java", "moe_experts": ["security", "data"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "37f8d38c037f5b6bb5c85ada1781b93745f6d001"} {"id": "train_049973", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: expert\nLanguage: Java\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["security", "data"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "ea335da809db4488cf67dd5f93f5ea4bb66def04"} {"id": "train_049974", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["research", "data"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "5438d98ec27ff43a6af5907c6c31ecad162a94ea"} {"id": "train_049975", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "SQL", "moe_experts": ["data", "evaluation"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "16f4181870f6c40f9570e557e2d8c7ca2eaa9c4c"} {"id": "train_049976", "created": "2026-01-01", "topic": "Dataset curation pipelines and quality attribution", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Dataset curation pipelines and quality attribution\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["performance", "governance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "9a67cc5c6ed44687337fe2c006e9f2f7ddfbef14"} {"id": "train_049977", "created": "2026-01-01", "topic": "Latency, cost, throughput optimization", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Latency, cost, throughput optimization\nDifficulty: advanced\nLanguage: TypeScript\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["security", "research"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "0c6bcf63c3a95692146022677efd92368d1ad5ae"} {"id": "train_049978", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: Java\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "Java", "moe_experts": ["performance", "security"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "badbd52cb6670b3179120f78d4ace323117d1516"} {"id": "train_049979", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "self_grade", "difficulty": "expert", "instruction": "Task: self_grade\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 7,\n \"confidence\": 0.7,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Bash", "moe_experts": ["security", "evaluation"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "334fea70158f94cc70f23a8f2618b88fa96eb550"} {"id": "train_049980", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Governance, provenance, and audit logging\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nMetrics:\n- pass@k\n- time-to-green\n- cost-per-fix\n- regression rate\n", "metadata": {"language": "Bash", "moe_experts": ["security", "data"], "cost_budget": 5, "latency_ms_target": 50, "audit_required": true}, "hash": "e5e419c270772c329026fea279174fcf97437cb2"} {"id": "train_049981", "created": "2026-01-01", "topic": "Self-grading and automated model critique", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Self-grading and automated model critique\nDifficulty: advanced\nLanguage: Rust\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 8,\n \"confidence\": 0.8,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "Rust", "moe_experts": ["data", "security"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "b18f63d1f6d99efc521b31d56d26be3b66efebeb"} {"id": "train_049982", "created": "2026-01-01", "topic": "SWE-bench real-repo evaluation and leakage control", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: SWE-bench real-repo evaluation and leakage control\nDifficulty: advanced\nLanguage: Bash\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["agent", "coding"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "2de24f3018c2218734bca261c99a7f5bbc3af0e3"} {"id": "train_049983", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: SQL\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "SQL", "moe_experts": ["evaluation", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "5ae4d946b284ef00b32237a636fa751c5f19cb17"} {"id": "train_049984", "created": "2026-01-01", "topic": "Failure engineering and resilience testing", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Failure engineering and resilience testing\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["security", "agent"], "cost_budget": 4, "latency_ms_target": 100, "audit_required": true}, "hash": "dda724f4f3acb37a8e025b35841baad288a9b25a"} {"id": "train_049985", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: JavaScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nPatch:\n```diff\n- result = a / b\n+ if b == 0:\n+ raise ValueError('division by zero')\n+ result = a / b\n\n```", "metadata": {"language": "JavaScript", "moe_experts": ["research", "data"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "2cfd709b5f0b7878635daedd111e50fa63a2bc2a"} {"id": "train_049986", "created": "2026-01-01", "topic": "Governance, provenance, and audit logging", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and audit logging\nDifficulty: expert\nLanguage: TypeScript\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "TypeScript", "moe_experts": ["evaluation", "research"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "a582e6bb4ce2b542d8cfb76b45a96d001b14ecab"} {"id": "train_049987", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nLanguage: Python\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Python", "moe_experts": ["evaluation", "research"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "b5ba44862c5a40ea16f0bd14f106ea3e07bee0ab"} {"id": "train_049988", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nFailure: regression detected\nCause: missing edge case\nFix: add guard + test\n", "metadata": {"language": "C#", "moe_experts": ["performance", "governance"], "cost_budget": 2, "latency_ms_target": 50, "audit_required": true}, "hash": "162a002e3a3372573ff822b4c87bd7c90d86a54f"} {"id": "train_049989", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Go\nContext: Research benchmark validation environment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "performance"], "cost_budget": 2, "latency_ms_target": 200, "audit_required": true}, "hash": "e53100faa005b1bab41feb3e7baee78b98cf2aab"} {"id": "train_049990", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: expert\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["data", "coding"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "334dca623ce42a9f92eccb2e437026a5b43b84c3"} {"id": "train_049991", "created": "2026-01-01", "topic": "Model merging, distillation, and continual learning", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continual learning\nDifficulty: expert\nLanguage: Bash\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["security", "governance"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "3a040a15f2d024244458de24a7ae966a0eb451d9"} {"id": "train_049992", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "tool_trace", "difficulty": "advanced", "instruction": "Task: tool_trace\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: TypeScript\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n[\n {\n \"tool\": \"repo_checkout\",\n \"args\": {\n \"commit\": \"abc123\"\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"unit\"\n }\n },\n {\n \"tool\": \"apply_patch\",\n \"args\": {\n \"files\": 1\n }\n },\n {\n \"tool\": \"run_tests\",\n \"args\": {\n \"suite\": \"full\"\n }\n }\n]", "metadata": {"language": "TypeScript", "moe_experts": ["data", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "0e41438e2fdc8cce8dff5bad1f66a0c7b8aa6056"} {"id": "train_049993", "created": "2026-01-01", "topic": "Extended context, memory, and repo graphs", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Extended context, memory, and repo graphs\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "data"], "cost_budget": 5, "latency_ms_target": 200, "audit_required": true}, "hash": "64639bd9d60c85378330418071d5291722392384"} {"id": "train_049994", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nLanguage: Bash\nContext: Offline air-gapped deployment\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Bash", "moe_experts": ["evaluation", "agent"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "8dc6d929663932b75372b541e444d150a4b2f48d"} {"id": "train_049995", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: Go\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Go", "moe_experts": ["evaluation", "agent"], "cost_budget": 2, "latency_ms_target": 100, "audit_required": true}, "hash": "f316bb570498739fb286c4553cc0b23afafbb4ce"} {"id": "train_049996", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "self_grade", "difficulty": "advanced", "instruction": "Task: self_grade\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nLanguage: SQL\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\n{\n \"score\": 9,\n \"confidence\": 0.9,\n \"notes\": \"Passed tests; minimal diff; no security flags.\"\n}", "metadata": {"language": "SQL", "moe_experts": ["agent", "performance"], "cost_budget": 4, "latency_ms_target": 200, "audit_required": true}, "hash": "f7c348b97f6d5532d620541df2bb7f76a0b28dc9"} {"id": "train_049997", "created": "2026-01-01", "topic": "Secure code generation, red-teaming, and policy gates", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Secure code generation, red-teaming, and policy gates\nDifficulty: advanced\nLanguage: C#\nContext: Latency-critical production API\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "C#", "moe_experts": ["performance", "research"], "cost_budget": 3, "latency_ms_target": 100, "audit_required": true}, "hash": "e42d687bffefb200c01de83abbba7c976ec22778"} {"id": "train_049998", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) routing and sparsity", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) routing and sparsity\nDifficulty: advanced\nLanguage: Rust\nContext: Regulated enterprise with audit requirements\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "Rust", "moe_experts": ["coding", "governance"], "cost_budget": 5, "latency_ms_target": 100, "audit_required": true}, "hash": "21f57f818b82078b7b9066b97207d213d98edb27"} {"id": "train_049999", "created": "2026-01-01", "topic": "Tool calling protocols and sandboxed execution", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Tool calling protocols and sandboxed execution\nDifficulty: expert\nLanguage: TypeScript\nContext: Large monorepo with flaky tests and strict CI\nProduce frontier-grade, production-valid output.", "input": "", "output": "Facts:\n- Modern AI coding success is measured by test pass rates, iteration cost, and reliability.\n- Agentic and multi-agent systems outperform single-pass generation.\n\nExpert guidance with risks, metrics, and governance controls.", "metadata": {"language": "TypeScript", "moe_experts": ["coding", "governance"], "cost_budget": 4, "latency_ms_target": 50, "audit_required": true}, "hash": "14e0c89382f8b94ac4e3af5b4f12da8d6d6923d9"}