openral-wam

OpenRAL World Action Model (WAM) layer — WorldModel Protocol + Rollout schema + NullWorldModel stub.

Scaffold status (2026-05-18). This package ships the Protocol surface only: WorldModel, Rollout, and a NullWorldModel stub for plumbing tests. Concrete adapters (Cosmos Predict, UnifoLM-WMA-0, IRASim) land in v0.3+ per CLAUDE.md §6.3. Those concrete generative WAM adapters ship as separate downstream packages in the private OpenRAL Pro monorepo — this public package stays the Protocol/contract surface.

Layer

CLAUDE.md §6.1 Layer 5 — generative simulator used by the planning loop for mental simulation (gating action chunks), failure anticipation, and replanning.

Three integration patterns (CLAUDE.md §6.3)

Mental simulation (gating) — sample N short rollouts before committing an action chunk. Threshold via Rollout.confidence.
Failure anticipation — continuous predicted-vs-observed discrepancy detector. The WorldStateAggregator and the WAM share a trace span; divergence raises a FailureTrigger.
Replanning loop — propose alternative subgoals as visual prompts; the reasoner re-prompts with the predicted-but-failed trajectory as context.

All three consume the same surface — WorldModel.rollout(...) → Rollout — so the Protocol is sufficient for v0.2.

Public surface

from openral_wam import WorldModel, Rollout, NullWorldModel

WorldModel — structural Protocol every WAM adapter satisfies.
Rollout — Pydantic v2 schema for the predicted trajectory.
NullWorldModel — identity stub returning the input state for horizon steps.

ADRs

Pydantic v2 preferred over @dataclass for schemas/configs.
CLAUDE.md §6.3 — the canonical write-up of the three integration patterns; a dedicated ADR will follow when the first concrete adapter lands.

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

openral_wam-0.3.0.tar.gz (7.5 kB view details)

Uploaded Aug 8, 2026 Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

The dropdown lists show the available interpreters, ABIs, and platforms. Enable javascript to be able to filter the list of wheel files.

openral_wam-0.3.0-py3-none-any.whl (6.9 kB view details)

Uploaded Aug 8, 2026 Python 3

File details

Details for the file openral_wam-0.3.0.tar.gz.

File metadata

Download URL: openral_wam-0.3.0.tar.gz
Upload date: Aug 8, 2026
Size: 7.5 kB
Tags: Source
Uploaded using Trusted Publishing? Yes
Uploaded via: twine/7.0.0 CPython/3.13.14

File hashes

Hashes for openral_wam-0.3.0.tar.gz
Algorithm	Hash digest
SHA256	`242361e109818cd0626acdee5bc641877b5ce527d75b6c5ab6938b989eaf2edb`
MD5	`cb7ce213867d721c7f92973605ade5b2`
BLAKE2b-256	`48ead0306075991e4576807df4cd2f917c27fe5ac1aa9bba70cf0cf6cab6408e`

See more details on using hashes here.

Provenance

The following attestation bundles were made for openral_wam-0.3.0.tar.gz:

Publisher: release-pypi.yml on OpenRAL/openral

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

Statement:
- Statement type: https://in-toto.io/Statement/v1
- Predicate type: https://docs.pypi.org/attestations/publish/v1
- Subject name: openral_wam-0.3.0.tar.gz
- Subject digest: 242361e109818cd0626acdee5bc641877b5ce527d75b6c5ab6938b989eaf2edb
- Sigstore transparency entry: 2385970939
- Sigstore integration time: Aug 8, 2026
Source repository:
- Permalink: OpenRAL/openral@be4417627a78f7b3151491ff1e33b41cc51019a4
- Branch / Tag: refs/tags/v0.3.0
- Owner: https://github.com/OpenRAL
- Access: public
Publication detail:
- Token Issuer: https://token.actions.githubusercontent.com
- Runner Environment: github-hosted
- Publication workflow: release-pypi.yml@be4417627a78f7b3151491ff1e33b41cc51019a4
- Trigger Event: push