earakely-scale/WordPressEnv-REA

wordpress

wordpress

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

Tables

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

Run it

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

WordPress on AgentEnv, from a REA capture

What this is. A replay environment of WordPress 7.1 (GPL-2.0-or-later), built from one recorded browser session against a self-hosted instance seeded with synthetic data. REA projected the capture into an environment manifest (21 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 WordPress. Tasks are graded deterministically, with no judge.

What's in it

• 21 tools (17 GET, 2 POST, 1 DELETE, 1 PUT) over 64 recorded exchanges on http://127.0.0.1:3504; no recorded group excluded.
• Path parameters: 8 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. With the four format rules alone the manifest has 21 tools.
• Unknowns the manifest states explicitly: authorization model, response semantics beyond recorded exchanges, state transitions between tools.
• 12 retrieval tasks (evals/wordpress-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.
• 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_5cf0a94d1d064315ce4c808f2d6d52759521633933272f25f2fcabee43622ba5 (scoped HAR sha256 f00e380bba7558ce78174d8ee8458f4db2684782125e67436ca37d88501cf82c, 241,446 bytes), manifest view ev_df7132f9f74e38314dd0e5deaf1d13801860a9a7f7d622639495b030a01c1388, confidence inferred, authority historical-reference. The REA build is rea-agents 6.7.0 from main at commit 3fca2e0c, which merged the environment commands (morluto/rea#4954, #4955, #4975, #4999 and #5182); 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_wp_json_wp_users_by_id	get_wp_json_wp_users_by_id	id=2	slug	ada-finch
qa-02-get_wp_json_wp_pages_by_id	get_wp_json_wp_pages_by_id	id=16	slug	about-the-gazette
qa-03-get_wp_json_wp_tags_by_id	get_wp_json_wp_tags_by_id	id=6	description	Posts that quote the spring and neap tide tables
qa-04-get_wp_json_wp_categories_by_id	get_wp_json_wp_categories_by_id	id=2	description	Quarry counts, stone audits and the ledger reference book
qa-05-get_wp_json_wp_posts_by_id	get_wp_json_wp_posts_by_id	id=4	slug	granite-ledger-quarterly-stone-count
qa-06-get_wp_json_wp_comments_by_id	get_wp_json_wp_comments_by_id	id=2	author_name	Copperfield Bindery
qa-07-get_wp_json_wp_search	get_wp_json_wp_search	search=lamp	[0].title	Visiting the lamp house
qa-08-get_wp_json_wp_pages	get_wp_json_wp_pages	no arguments	[0].slug	visiting-the-lamp-house
qa-09-get_wp_json_wp_users	get_wp_json_wp_users	no arguments	[0].name	Ada Finch
qa-10-get_wp_json_wp_tags	get_wp_json_wp_tags	no arguments	[0].slug	cold-chain
qa-11-get_wp_json_wp_categories	get_wp_json_wp_categories	no arguments	[0].name	Blue Heron Kitchen
qa-12-get_wp_json_wp_posts	get_wp_json_wp_posts	per_page=20	[0].slug	heron-wharf-bread-oven

Episodes

model	tasks	passed	pass rate	mean tool calls	mean input tokens
openai/gpt-oss-120b:groq	12	12	100%	1.0	7,804
zai-org/GLM-5.3-Flash:zai-org	12	12	100%	1.0	9,279

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/WordPressEnv-REA --bundle wordpress --task smoke          # Docker, no model
agent-env hf run earakely-scale/WordPressEnv-REA --bundle wordpress --eval wordpress-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. WordPress ran in Docker on a throwaway VM, bound to loopback, seeded through its own API with synthetic records (site: Lighthouse Rota Gazette (tagline: Dispatches from the lamp house, the quarry ledger and the tide-table kitchen); users: Ada Finch (author), Rui Castellanos (editor), Mina Okafor (author), Tomas Vey (author); categories: Granite Ledger, Blue Heron Kitchen, Saltmarsh Depot, Copperfield Annex; tags: tide tables, night shift, lamp oil, cold chain, stone count; posts: Granite Ledger closes the quarter at 412 blocks, Seaward stack re-measured after the spring storm, Blue Heron Kitchen opens its tide-table menu, Smoked kelp broth for the lamp house night shift, Winter lamp oil budget set at...); 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	WordPress (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)	21 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 WordPress; 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 WordPress instance over synthetic data; WordPress is GPL-2.0-or-later software by the WordPress contributors 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.