earakely-scale/RedpandaConsoleEnv-REA

redpanda-console

redpanda-console

An AgentEnv bundle with its recorded runs, published with agentenv-hf.

Tables

• redpanda-console_tasks: one row per task of the bundle, with its prompt, envs, verifiers and steps.
• redpanda-console_episodes: one row per recorded run, with its status, model, reward and scores, the prompt and response, and the agent's transcript as chat messages when its trajectory format is one agentenv-hf reads.

raw/redpanda-console.jsonl holds each run's record and native trajectory, and bundles/redpanda-console/ the bundle itself.

Run it

pip install agentenv-framework huggingface_hub
hf download earakely-scale/RedpandaConsoleEnv-REA --repo-type dataset --local-dir agentenv-dataset
agent-env run agentenv-dataset/bundles/redpanda-console

Redpanda Console on AgentEnv, from a REA capture

What this is. A replay environment of Redpanda Console 3.12.0 (BSL-1.1), built from one recorded browser session against a self-hosted instance seeded with synthetic data. REA projected the capture into an environment manifest (37 tools), and rea environment-export wrote the bundle you can run from this dataset. Every tool call replays a retained record from the capture, cites the record and the Evidence it came from, and nothing executes Redpanda Console. Tasks are graded deterministically, with no judge.

What's in it

• 37 tools (31 GET, 3 POST, 2 DELETE, 1 PATCH) over 87 recorded exchanges on http://127.0.0.1:3305; no recorded group excluded.
• Path parameters: 6 numeric, 5 sibling-variance. REA's four format rules (numeric, uuid, hex, iso-date) applied unconditionally; the three opt-in rules (after-uid-literal, opaque-token, sibling-variance) were requested and are labelled per parameter. With the four format rules alone the manifest has 54 tools.
• Unknowns the manifest states explicitly: authorization model, response semantics beyond recorded exchanges, state transitions between tools.
• 8 retrieval tasks (evals/redpanda-console-retrieval.toml): each asks one factual question about a seeded object, with a distinct answer that the app does not ship by default, answerable through a recorded exchange; reward 1 if the response contains the recorded value, else 0. Plus smoke, which deploys the environment and checks replay through the MCP endpoint.
• 16 recorded episodes by openai/gpt-oss-120b:groq, zai-org/GLM-5.3-Flash:zai-org through the Hugging Face router, with their chat transcripts.
• Provenance: capture ev_54e03e32f9f70689e0cde3c4a220fa8f93b4b0948cd7f4cc4db8f59c40324382 (scoped HAR sha256 5bc50a249c91ea871c1f1cf5d93990674c4436ebde14eca474ae01454715a5b2, 197,451 bytes), manifest view ev_60042ee4b8db9701a8b70151626c426c9bb659811a29fc3c6d596f76d1582874, confidence inferred, authority historical-reference. The REA build is a prerelease of the environment commands proposed upstream in morluto/rea#4954, #4955, #4975 and #4999; the bundle installs it from the tarball it ships, so the environment runs without waiting for an npm release.

What replay means, and does not

A tool call returns the recorded response whose path and query values best match the arguments, with replay.exact_match telling you whether it matched exactly or fell back to the closest record. Reads of seeded objects behave like the application. Writes return the recorded answer of the one write the session made, and the environment has no state: a created object does not appear in a later list. This is what makes retrieval tasks hard-gradable and write tasks weak; the manifest's unknowns say so.

Tasks and grading

task	tool	arguments	field	answer
qa-01-get_schema_registry_schemas_ids_versions	get_schema_registry_schemas_ids_versions_by_id	id=6	[0].subject	saltmarsh-depot.crate-manifests-value
qa-02-get_schema_registry_subjects_versions_la	get_schema_registry_subjects_versions_latest_by_subject_id	subject_id=lighthouse-rota.shift-handover-value	schemas[0].references[0].name	com.example.rota.Keeper
qa-03-get_schema_registry_subjects_lighthouse_	get_schema_registry_subjects_lighthouse_rota_versions_by_id	id=1	name	lighthouse-rota.keeper
qa-04-get_brokers_config_by_id	get_brokers_config_by_id	id=0	brokerConfigs[0].value	internal://0.0.0.0:9092
qa-05-get_schema_registry_subjects_granite_led	get_schema_registry_subjects_granite_ledger_versions_by_id	id=2	name	granite-ledger.stone-counts-value
qa-06-get_schema_registry_subjects_lighthouse_	get_schema_registry_subjects_lighthouse_rota_versions_refere	id=1	[0].usages[0].subject	lighthouse-rota.shift-handover-value
qa-07-get_topics_granite_ledger_configuration	get_topics_granite_ledger_configuration	no arguments	topicDescription.configEntries[0].name	compression.type
qa-08-get_schema_registry_subjects	get_schema_registry_subjects	no arguments	[0].name	blue-heron-kitchen.tide-orders-value

Episodes

model	tasks	passed	pass rate	mean tool calls	mean input tokens
openai/gpt-oss-120b:groq	8	8	100%	1.0	9,494
zai-org/GLM-5.3-Flash:zai-org	8	8	100%	1.1	11,363

Every episode is in the episodes config with its transcript; rewards are the deterministic verifier scores.

Every recorded episode passed.

Run it yourself

uv tool install agentenv-framework --with 'agentenv-hf @ git+https://github.com/earakely-scale/agentenv-hf-plugin@v0.3.0'
agent-env hf run earakely-scale/RedpandaConsoleEnv-REA --bundle redpanda-console --task smoke          # Docker, no model
agent-env hf run earakely-scale/RedpandaConsoleEnv-REA --bundle redpanda-console --eval redpanda-console-retrieval --model openai/gpt-oss-120b:groq

The retrieval eval needs a model endpoint: set [model] base_url and api_key in .agentenv/config.toml (the Hugging Face router works with any hf_ token). The bundle's agents/replay-qa is a small OpenAI-compatible tool-calling agent; any A2A agent can replace it.

How it was made

1. Redpanda Console ran in Docker on a throwaway VM, bound to loopback, seeded through its own API with synthetic records (five topics (granite-ledger.stone-counts, blue-heron-kitchen.tide-orders, lighthouse-rota.shift-handover, copperfield-annex.lamp-oil-deliveries, saltmarsh-depot.crate-manifests) holding 22 invented JSON messages, four consumer groups with committed offsets (two with lag), and five schema-registry subjects (Avro with a cross-subject reference, JSON Schema, Protobuf)); people and projects are invented.
2. Playwright drove a login, the main pages and a scripted read, write, read sequence through the app's JSON API; the HAR was scoped to the API, the login exchange dropped, secret-looking body fields scrubbed, headers and cookies removed.
3. rea inspect-web-network-capture retained the exchanges as Evidence; rea inspect-analysis-view projected the environment-manifest view; rea environment-export wrote this bundle; gen_tasks.py derived the retrieval tasks from the manifest and the recorded bodies; agent-env run recorded the episodes; agent-env hf publish wrote this dataset. Scripts: https://github.com/earakely-scale/rea-environments.

Compared with PortSimEnv on AgentEnv

	PortSimEnv-AgentEnv	Redpanda Console (this dataset)
Environment	A simulation: a container quay, ships, cranes, weather, with A2A ship agents	A replay of a real application's API, captured from one session
Observation	Tool results computed by the simulator as the week unfolds	Recorded responses, cited to Evidence
Action space	Planning tools (check_plan, confirm_berths, advance, message_agent)	37 tools mirroring the application's endpoints
State	Full: a virtual clock, triggers, frozen windows, agents that answer	None: every call is independent; writes replay one recorded answer
Reward	Deterministic, against a CP-SAT optimum or a re-planner anchor; continuous 0 to 1	Deterministic, recorded-value match; 0 or 1 per task
What it measures	Planning under uncertainty and negotiation	Tool discovery, parameter inference and faithful retrieval in a real API
Data provenance	Real 2024 port calls and weather; simulated workloads	Synthetic seed data on a self-hosted Redpanda Console; nothing real
How it was built	Hand-written plugin, months of work	Generated from a capture in one scripted pass
Where it is strong	Long-horizon decisions with hard rewards	Breadth: any app you can capture becomes an environment

The two are complementary: PortSimEnv shows what a hand-built simulation can grade; this dataset shows how far a capture alone gets you, and the manifest's unknowns list exactly what a stateful version would still have to specify.

Licence and credits

Bundle code and tasks: Apache-2.0. Recorded content: responses generated by a self-hosted Redpanda Console instance over synthetic data; Redpanda Console is BSL-1.1 software by Redpanda Data, Inc. and is not redistributed here beyond the recorded JSON responses. REA is MIT software by morluto; AgentEnv Framework is Apache-2.0 by Scale AI.