earakely-scale/N8nEnv-REA

n8n

n8n

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

Tables

• n8n_tasks: one row per task of the bundle, with its prompt, envs, verifiers and steps.
• n8n_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/n8n.jsonl holds each run's record and native trajectory, and bundles/n8n/ the bundle itself.

Run it

pip install agentenv-framework huggingface_hub
hf download earakely-scale/N8nEnv-REA --repo-type dataset --local-dir agentenv-dataset
agent-env run agentenv-dataset/bundles/n8n

n8n on AgentEnv, from a REA capture

What this is. An environment of n8n 1.123.85 (Sustainable Use License 1.0 (fair-code; .ee files under the n8n Enterprise License)), built from one recorded browser session against a self-hosted instance seeded with synthetic data. REA projected the capture into an environment manifest (33 tools) and an environment state, the entities the capture's reads showed, and rea environment-export wrote the bundle you can run from this dataset. Reads the state can render are answered from it, write tools whose effect the capture confirmed change it, and data/reset restores it. Nothing executes n8n. Tasks are graded deterministically, with no judge.

v0.2.0 rebuilds this environment from the same capture with a REA build that serves a state and confirmed writes; the tag v0.1.0 keeps the earlier replay-only version.

What's in it

• 33 tools (26 GET, 3 POST, 2 DELETE, 2 PATCH) over 72 recorded exchanges on http://127.0.0.1:3218; no recorded group excluded. 27 are offered; the other 6 are write tools the environment cannot apply (post_rest_tags, delete_rest_tags_lvqnhb0heimbid9r, patch_rest_tags_lvqnhb0heimbid9r, post_rest_workflows, patch_rest_workflows_by_workflow_id, post_rest_workflows_archive_by_workflow_id), left out rather than faked.
• Path parameters: 4 prefix-variance, 3 sibling-variance, 1 numeric. REA's four format rules (numeric, uuid, hex, iso-date) applied unconditionally; the four opt-in rules (after-uid-literal, opaque-token, sibling-variance, prefix-variance) were requested and are labelled per parameter. With the four format rules alone the manifest has 40 tools.
• State: 15 entities of 5 types. 17 read tools answer from it and 9 replay their recorded responses. Served write tools: delete_rest_workflows_by_workflow_id (delete).
• Unknowns the manifest states explicitly: authorization model, response semantics beyond recorded exchanges, state transitions between tools.
• 9 retrieval tasks (evals/n8n-retrieval.toml): each asks one factual question about a seeded object, with a distinct answer that the app does not ship by default; reward 1 if the response contains the recorded value, else 0. Plus smoke, which deploys the environment and checks it through the MCP endpoint.
• 2 write tasks (evals/n8n-writes.toml): each asks for a change in plain language and is graded from the environment's state through data/get, not from the agent's reply; reward 1 only if every check passes. Their grade is in scores.state; reward holds the retrieval verifier's score, so it is empty for them, as for smoke.
• 18 retrieval episodes and 4 write episodes, plus the smoke run, 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_2df554f077a7336e5d535fa9e49df0b2d984c88dfc06c774e0343f26f9176a3e (scoped HAR sha256 ad32e5104208daeb3488f6feefce0f189fe6ade22e3857907a10213009a022a3, 189,640 bytes), manifest view ev_f6b50ab95fe5bb1ed0499b4674785987415bcbd5dc473fdfb5b1b5e3b99d7fd8, confidence inferred, authority historical-reference. The REA build is rea-agents 6.7.0 from main 1629266e with open pull requests merged: morluto/rea#5600 at 370b9bef (which carries #5351 and #5352) and #5664 at 33b0643e. The bundle installs it from the tarball it ships, so the environment runs without waiting for an npm release.

What the state means, and does not

Reads the state can render, item, list and document views of the entities the capture's responses showed, answer from it. Other reads replay the recorded response whose path and query values best match the arguments, with replay.exact_match saying whether it matched exactly; a replayed read does not change after a write. A served write was confirmed by the capture itself: replaying the session against the extracted state, a later recorded read reproduced its effect and none contradicted it. A create generates ids and clock values deterministically, so the same calls after a reset give the same state; an update changes the members the call sends; a delete removes the entity. Each takes only the body members the recorded call showed and refuses others with a message that names the accepted ones. Its effect shows in the reads the state renders, except that a list the capture saw only filtered or in part keeps its recorded members: it drops a deleted entity but does not add a new one. The data plane at /agentenv takes data/reset (restore the extracted state), data/add and data/get (return it), which is how write tasks are graded.

Not modelled: side effects a write has beyond its own entity (counters, copies in other collections, related entities), preconditions the application enforces before a write, new entities in lists whose recorded calls were filtered, and anything the capture never showed. The manifest's unknowns list what a capture cannot establish. Replayed against a live n8n 1.123.85 seeded by the same script, listing, reading and deleting an archived workflow agreed apart from ids the seed generates. n8n refuses to delete a workflow that is not archived while the environment deletes it, the workflow list replays its recording so a delete does not show in it, and archiving is not offered.

Tasks and grading

Retrieval

task	tool	arguments	field	answer
qa-01-get_rest_workflows_by_workflow_id	get_rest_workflows_by_workflow_id	workflow_id=FkMwxmXNGaGVsXFG	data.description	Totals the quarterly block count for the Granite Ledger and names the auditor.
qa-02-get_rest_workflows_executions_last_succe	get_rest_workflows_executions_last_successful_by_workflow_id	workflow_id=FkMwxmXNGaGVsXFG	data.workflowData.name	Granite Ledger stone count digest
qa-03-get_rest_projects_pliwgg1jdk2kbkgs_data_	get_rest_projects_pliwgg1jdk2kbkgs_data_tables_rows_by_data	data_table_id=6s4t9bREtisTcoCX, skip=0, take=20	data.data[0].signed_by	Ada Finch
qa-04-get_rest_data_tables_global	get_rest_data_tables_global	skip=0, take=20	data.data[0].name	Quarry Block Tally
qa-05-get_rest_executions	get_rest_executions	limit=10	data.results[0].workflowName	Blue Heron Kitchen sitting planner
qa-06-get_rest_workflows	get_rest_workflows	sortBy=name:asc, skip=0, take=50	data[0].description	Plans each tide-table sitting and decides on the second dining room.
qa-07-get_rest_workflows_by_workflow_id	get_rest_workflows_by_workflow_id	workflow_id=FkMwxmXNGaGVsXFG	data.nodes[0].name	Start stone count
qa-08-get_rest_projects_pliwgg1jdk2kbkgs_data_	get_rest_projects_pliwgg1jdk2kbkgs_data_tables_rows_by_data	data_table_id=6s4t9bREtisTcoCX, skip=0, take=20	data.data[0].barge	Pale Meridian
qa-09-get_rest_workflows	get_rest_workflows	sortBy=name:asc, skip=0, take=50	data[0].tags[0].name	Blue Heron Kitchen

Writes

Graded from the environment's state through data/get (scores.state in the episodes table; reward 1 only if every check passes).

task	prompt	checks
write-delete	In this n8n instance, delete the workflow "Granite Ledger stone count digest", and only that workflow.	the Granite Ledger workflow is gone; the other three workflows remain
write-delete-two	In this n8n instance, the Saltmarsh Depot and Blue Heron Kitchen workflows are retired. Delete both of them and nothing else.	the Saltmarsh Depot workflow is gone; the Blue Heron Kitchen workflow is gone; the other two workflows remain

Episodes

model	retrieval episodes	passed	pass rate	mean tool calls	mean input tokens
openai/gpt-oss-120b:groq	9	9	100%	1.0	10,751
zai-org/GLM-5.3-Flash:zai-org	9	9	100%	1.0	12,695

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

model	write episodes	passed
openai/gpt-oss-120b:groq	2	2
zai-org/GLM-5.3-Flash:zai-org	2	2

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/N8nEnv-REA --bundle n8n --task smoke          # Docker, no model
agent-env hf run earakely-scale/N8nEnv-REA --bundle n8n --eval n8n-retrieval --model openai/gpt-oss-120b:groq
agent-env hf run earakely-scale/N8nEnv-REA --bundle n8n --eval n8n-writes --model openai/gpt-oss-120b:groq

The evals need 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. n8n ran in Docker on a throwaway VM, bound to loopback, seeded through its own API with synthetic records (owner Rea Admin (rea-admin@example.test); 5 tags (Granite Ledger, Blue Heron Kitchen, Lighthouse Rota, Saltmarsh Depot, Quarterly Audit); 4 workflows built from core nodes only, no credentials (Granite Ledger stone count digest, Blue Heron Kitchen sitting planner, Lighthouse Rota shift handover, Saltmarsh Depot barge drop log); 2 manual executions; 2 data tables (Lamp Oil Drops with 4 rows, Quarry Block Tally with 3 rows)); 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 and the state it serves; gen_tasks.py derived the retrieval tasks from the manifest and the recorded bodies; the write tasks were written by hand against the served write tools and the seeded names; 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	n8n (this dataset)
Environment	A simulation: a container quay, ships, cranes, weather, with A2A ship agents	An application's API rebuilt from one captured session
Observation	Tool results computed by the simulator as the week unfolds	Answers rendered from the extracted state, or recorded responses, cited to Evidence
Action space	Planning tools (check_plan, confirm_berths, advance, message_agent)	27 tools mirroring the application's endpoints
State	Full: a virtual clock, triggers, frozen windows, agents that answer	A state extracted from the capture (15 entities); served writes change it; data/reset restores it
Reward	Deterministic, against a CP-SAT optimum or a re-planner anchor; continuous 0 to 1	Deterministic: recorded-value match for retrieval, state checks for writes; 0 or 1 per task
What it measures	Planning under uncertainty and negotiation	Tool discovery, parameter inference and faithful retrieval in a real API, and changing its state
Data provenance	Real 2024 port calls and weather; simulated workloads	Synthetic seed data on a self-hosted n8n; 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 and the state's limits above say what it still leaves out.

Licence and credits

Bundle code and tasks: Apache-2.0. Recorded content: responses generated by a self-hosted n8n instance over synthetic data; n8n is Sustainable Use License 1.0 (fair-code; .ee files under the n8n Enterprise License) software by n8n GmbH 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.