earakely-scale/NocoDBEnv-REA

nocodb

nocodb

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

Tables

• nocodb_tasks: one row per task of the bundle, with its prompt, envs, verifiers and steps.
• nocodb_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/nocodb.jsonl holds each run's record and native trajectory, and bundles/nocodb/ the bundle itself.

Run it

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

NocoDB on AgentEnv, from a REA capture

What this is. A replay environment of NocoDB 2026.09.1 (AGPL-3.0), built from one recorded browser session against a self-hosted instance seeded with synthetic data. REA projected the capture into an environment manifest (28 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 NocoDB. Tasks are graded deterministically, with no judge.

What's in it

• 28 tools (23 GET, 3 POST, 1 DELETE, 1 PATCH) over 105 recorded exchanges on http://127.0.0.1:3103; 0 recorded groups excluded as UI pages, assets or other origins.
• Path parameters: 24 opaque-token, 1 numeric. 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. Without them the manifest has 35 tools, one per literal path.
• Unknowns the manifest states explicitly: authorization model, response semantics beyond recorded exchanges, state transitions between tools.
• 12 retrieval tasks (evals/nocodb-retrieval.toml): each asks one factual question answerable through exactly one 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.
• 24 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_c2296ec7318f8978720a75499d6fb34a6c5f5f8943f2da32acc1eaaad450e08e (scoped HAR sha256 fe7e31fddd225d7bd00ecd8739fd0a2a474ef7a36aa79398900bc81499093f65, 324,797 bytes), manifest view ev_12b6ec59bee2946b7b74a675d71335b5c56a2c026712583c6342ca0a752d8616, 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_db_meta_projects_tables_by_uid	get_db_meta_projects_tables_by_uid	uid=p6xibd8cmh0sl25, includeM2M=false	list[0].title	Vessels
qa-02-get_meta_tables_views_by_uid	get_meta_tables_views_by_uid	uid=m4eijvqyydf5ont	list[0].title	Vessels
qa-03-get_tables_records_by_uid_id	get_tables_records_by_uid_id	uid=m4eijvqyydf5ont, id=1	name	Granite Ledger
qa-04-get_meta_tables_by_uid	get_meta_tables_by_uid	uid=m4eijvqyydf5ont	columnsById.c1jvdd1hcqbv7om.title	nc_created_by
qa-05-get_db_meta_projects_users_by_uid	get_db_meta_projects_users_by_uid	uid=p6xibd8cmh0sl25	users.list[0].email	rea-admin@example.test
qa-06-get_tables_records_by_uid	get_tables_records_by_uid	uid=m4eijvqyydf5ont, limit=25	list[0].name	Granite Ledger
qa-07-get_meta_bases_tables_by_uid	get_meta_bases_tables_by_uid	uid=p6xibd8cmh0sl25	list[0].table_name	nc_sw86___vessels
qa-08-get_meta_bases_by_uid	get_meta_bases_by_uid	uid=p6xibd8cmh0sl25	title	Lighthouse Rota
qa-09-get_db_data_noco_views_by_uid_uid_2_uid_	get_db_data_noco_views_by_uid_uid_2_uid_3	uid=p6xibd8cmh0sl25, uid_2=m4eijvqyydf5ont, uid_3=vwde0b34abk8eh18, offset=0, limit=100, includeSortAndFilterColumns=true, where=, include_row_color=true, include_button_filter_columns=true	list[0].name	Granite Ledger
qa-10-get_internal_nc_by_uid	get_internal_nc_by_uid	uid=p6xibd8cmh0sl25, operation=baseSchema	name	Lighthouse Rota
qa-11-get_db_meta_projects_by_uid	get_db_meta_projects_by_uid	uid=p6xibd8cmh0sl25	description	Synthetic berth planning board for the Lighthouse Rota showcase
qa-12-get_meta_bases	get_meta_bases	no arguments	list[0].title	Getting Started

Episodes

model	tasks	passed	pass rate	mean tool calls	mean input tokens
openai/gpt-oss-120b:groq	12	12	100%	1.0	9,195
zai-org/GLM-5.3-Flash:zai-org	12	12	100%	1.0	11,416

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/NocoDBEnv-REA --bundle nocodb --task smoke          # Docker, no model
agent-env hf run earakely-scale/NocoDBEnv-REA --bundle nocodb --eval nocodb-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. NocoDB ran in Docker on a throwaway VM, bound to loopback, seeded through its own API with synthetic records ({'bases': 1, 'tables': 2, 'rows_per_table': 8, 'base_title': 'Lighthouse Rota', 'tables_list': ['Vessels', 'Port Calls']}); 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	NocoDB (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)	28 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 NocoDB; 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 NocoDB instance over synthetic data; NocoDB is AGPL-3.0 software by NocoDB 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.