earakely-scale/NomadEnv-REA

nomad

nomad

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

Tables

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

Run it

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

HashiCorp Nomad on AgentEnv, from a REA capture

What this is. A replay environment of HashiCorp Nomad 1.11.3 (BUSL-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 (39 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 HashiCorp Nomad. Tasks are graded deterministically, with no judge.

What's in it

• 39 tools (34 GET, 4 POST, 1 DELETE) over 131 recorded exchanges on http://127.0.0.1:3304; no recorded group excluded.
• Path parameters: 5 uuid, 1 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 46 tools.
• Unknowns the manifest states explicitly: authorization model, response semantics beyond recorded exchanges, state transitions between tools.
• 12 retrieval tasks (evals/nomad-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_09880f9a941052d3335d98097843a3ad3774778266aabd85fbe9093de61b6ad1 (scoped HAR sha256 61df40951c6ee1872176416df0d5b0a6c2f8eaea91230c7f01b16ea47343df83, 615,790 bytes), manifest view ev_12982eae5eeb6c118ce31026d3377f17232a0b872b74ef36620d7c418752688d, 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_node_by_uuid	get_node_by_uuid	uuid=976166e0-7b01-9d55-5a52-14f3b53d1ec9	Events[0].Message	Node registered
qa-02-get_job_by_job_id	get_job_by_job_id	job_id=blue-heron-kitchen-orders, namespace=lighthouse-rota	TaskGroups[0].Meta.sitting	second sitting at 21:30
qa-03-get_allocation_by_uuid	get_allocation_by_uuid	uuid=f8af374d-bd2c-81c3-9f09-24699e41b107	Name	granite-ledger-stone-count.quarry-tally[0]
qa-04-get_job_blue_heron_kitchen_orders_evalua	get_job_blue_heron_kitchen_orders_evaluations	namespace=lighthouse-rota	[0].TriggeredBy	job-register
qa-05-get_node_pool_saltmarsh_depot	get_node_pool_saltmarsh_depot	no arguments	Description	Cold-store hosts at the Saltmarsh Depot (no hosts enrolled yet)
qa-06-get_job_rea_scratch_tide_probe_versions	get_job_rea_scratch_tide_probe_versions	namespace=lighthouse-rota	Versions[0].Meta.purpose	scratch tide probe edited during the capture
qa-07-get_namespace_lighthouse_rota	get_namespace_lighthouse_rota	no arguments	Description	Night shifts, lamp oil and tide tables for the Lighthouse Rota
qa-08-get_job_blue_heron_kitchen_orders_versio	get_job_blue_heron_kitchen_orders_versions	namespace=lighthouse-rota	Versions[0].Meta.kitchen_menu	neap tide menu with oat crackers
qa-09-get_node_pool_saltmarsh_depot_jobs	get_node_pool_saltmarsh_depot_jobs	no arguments	[0].Name	saltmarsh-depot-cold-store
qa-10-get_allocations	get_allocations	namespace=*, resources=true, task_states=false	[0].Name	blue-heron-kitchen-orders.tide-table-counter[0]
qa-11-get_namespace_granite_ledger	get_namespace_granite_ledger	no arguments	Description	Quarry audits and stone counts for the Granite Ledger
qa-12-get_job_blue_heron_kitchen_orders_alloca	get_job_blue_heron_kitchen_orders_allocations	namespace=lighthouse-rota	[0].Name	blue-heron-kitchen-orders.tide-table-counter[1]

Episodes

model	tasks	passed	pass rate	mean tool calls	mean input tokens
openai/gpt-oss-120b:groq	12	12	100%	1.0	11,277
zai-org/GLM-5.3-Flash:zai-org	12	12	100%	1.0	13,998

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/NomadEnv-REA --bundle nomad --task smoke          # Docker, no model
agent-env hf run earakely-scale/NomadEnv-REA --bundle nomad --eval nomad-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. HashiCorp Nomad ran in Docker on a throwaway VM, bound to loopback, seeded through its own API with synthetic records (two namespaces, one node pool with no nodes, metadata on the dev node, six jobs across three namespaces (two running service jobs, a service job pending on the empty pool, a finished batch job, a periodic batch job, a parameterized batch job with one dispatched child) and a second version of one service job); 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	HashiCorp Nomad (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)	39 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 HashiCorp Nomad; 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 HashiCorp Nomad instance over synthetic data; HashiCorp Nomad is BUSL-1.1 software by HashiCorp 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.