Provider Configuration Index¶
This index is generated from the in-repo provider catalog, provider profiles, request policies, and runtime manifests. It is the operator-facing contract for which inputs enable each provider, which optional packages stay host-owned, which assets the host must preserve, and which command to run first when configuration fails.
Evidence levels:
scaffold: name reservation only; no real runtime capability is claimed.fixture-tested: deterministic in-repo behavior is covered by local tests and fixtures.prepared-host: WorldForge ships a runtime manifest and smoke command, but the host supplies credentials, endpoints, optional packages, checkpoints, or robot-specific assets.live-smoke: a prepared host has preserved a sanitized run manifest for the provider and capability in the Live Smoke Evidence Registry.
WorldForge never stores secrets, host-local endpoint values, checkpoint archives, or robot controller credentials in this index.
| Provider | Evidence level | Required inputs | Optional inputs | Optional packages | Prepared-host assets | Default timeouts | First diagnostic |
|---|---|---|---|---|---|---|---|
mock |
fixture-tested |
none | none | none | none | none | uv run worldforge provider health mock |
cosmos-policy |
prepared-host |
COSMOS_POLICY_BASE_URL |
COSMOS_POLICY_API_TOKEN, COSMOS_POLICY_TIMEOUT_SECONDS, COSMOS_POLICY_EMBODIMENT_TAG, COSMOS_POLICY_MODEL, COSMOS_POLICY_RETURN_ALL_QUERY_RESULTS, COSMOS_POLICY_ALLOW_LOCAL_BASE_URL, COSMOS_POLICY_ALLOWED_HOSTS |
nvidia/cosmos-policy server | Cosmos-Policy Docker/runtime environment ALOHA policy server model checkpoints observation builder action translator |
health 10s x3; request 600s x1; poll 30s x3; download 600s x3 | uv run worldforge provider health cosmos-policy |
leworldmodel |
prepared-host |
LEWORLDMODEL_POLICY or LEWM_POLICY |
STABLEWM_HOME, LEWORLDMODEL_CACHE_DIR, LEWORLDMODEL_DEVICE |
torch stable_worldmodel |
LeWorldModel checkpoint checkpoint cache task-shaped tensors |
none | uv run worldforge provider health leworldmodel |
gr00t |
prepared-host |
GROOT_POLICY_HOST |
GROOT_POLICY_PORT, GROOT_POLICY_TIMEOUT_MS, GROOT_POLICY_API_TOKEN, GROOT_POLICY_STRICT, GROOT_EMBODIMENT_TAG |
gr00t.policy.server_client msgpack numpy pyzmq |
GR00T policy server embodiment assets action translator |
none | uv run worldforge provider health gr00t |
lerobot |
prepared-host |
LEROBOT_POLICY_PATH or LEROBOT_POLICY |
LEROBOT_POLICY_TYPE, LEROBOT_DEVICE, LEROBOT_CACHE_DIR, LEROBOT_EMBODIMENT_TAG |
lerobot | policy checkpoint policy cache embodiment action translator |
none | uv run worldforge provider health lerobot |
jepa |
prepared-host |
JEPA_MODEL_NAME |
JEPA_DEVICE, JEPA_MODEL_PATH |
torch | JEPA-WMS checkpoint torch-hub cache task-shaped observation, goal, and action tensors |
none | uv run worldforge provider health jepa |
genie |
scaffold |
GENIE_API_KEY |
none | none | none | none | uv run worldforge provider health genie |
Prepared-Host Smoke Commands¶
| Provider | Smoke command | Credential gate | Runtime ownership |
|---|---|---|---|
mock |
not smoke-testable from WorldForge |
none | in-repo deterministic local provider |
cosmos-policy |
COSMOS_POLICY_BASE_URL=http://127.0.0.1:8777 COSMOS_POLICY_ALLOW_LOCAL_BASE_URL=1 uv run worldforge-smoke-cosmos-policy --policy-info-json /path/to/policy_info.json --translator /path/to/translator.py:translate_actions --allow-translator-code |
COSMOS_POLICY_API_TOKEN |
WorldForge validates /act request/response and planning composition; host owns Cosmos-Policy reachability/CUDA/runtime, ALOHA observation construction, and translation of raw 14D rows into executable Action objects |
leworldmodel |
scripts/lewm-real --checkpoint ~/.stable-wm/pusht/lewm_object.ckpt --device cpu |
none | host installs the official LeWM loading path (stable_worldmodel.policy.AutoCostModel), torch, and compatible checkpoints |
gr00t |
GROOT_POLICY_HOST=127.0.0.1 GROOT_POLICY_PORT=5555 uv run --with msgpack --with pyzmq --with numpy python scripts/smoke_gr00t_policy.py --health-only --run-manifest .worldforge/runs/gr00t-health/run_manifest.json |
GROOT_POLICY_API_TOKEN |
host runs or reaches an Isaac GR00T policy server |
lerobot |
scripts/smoke_lerobot_policy.py --policy-path <repo-or-checkpoint> --device cpu |
none | host installs LeRobot and compatible policy checkpoints |
jepa |
uv run --with torch worldforge-smoke-jepa-wms --model-name jepa_wm_pusht --device cpu |
none | host supplies torch, facebookresearch/jepa-wms runtime dependencies, and task preprocessing |
genie |
not smoke-testable from WorldForge |
GENIE_API_KEY |
capability-fail-closed reservation; Project Genie has no supported automation API contract |
The generated block is checked by:
When a provider changes its profile, environment variables, runtime manifest, smoke command, or request policy, regenerate this page and inspect the diff before publishing.