[jonahgabriel]

Summary

• Adds the exact logical model, served-model, and launch-profile bindings to the inert RSD provenance receipt.
• Pins the selected base/artifact revisions, manifest commitment, W4A4/FP8 precision, hardware capability, and vLLM compatibility facts as fail-closed literals.
• Adds substitution/absence coverage and registers the model’s explicit serialization disposition.

Scope

This is an execution-disabled, topology-neutral provenance contract only. It adds no adapter, resolver, endpoint, credential, deployment, runtime activation, canary, or hostile-CI-provider behavior.

Validation

• uv run pytest -q tests/unit/runtime/test_rsd_model_artifact_provenance.py tests/unit/models/test_model_serialization_roundtrip.py (74 passed)
• uv run mypy src/ --strict (2,956 source files)
• pre-commit run --all-files
• Normal commit and pre-push hooks (no bypass)

Evidence-Ticket: OMN-17871
Evidence-Source: OCC#8290

[github-actions]

⚠️ Hostile Reviewer — DEGRADED (informational)

Critical findings: 0
Major findings: 0
Total findings: 0
Models succeeded: glm-review

Note: Fewer than 2 reviewer models succeeded. Degraded results are informational (OMN-8468/OMN-8524) and do not block merge. Error: cli_review exit 2 (fewer than 2 models succeeded — partial/total outage)

---

Semantics (OMN-17492 — the model finds, thread resolution gates)

Surface	Meaning	Blocks merge?
Review threads	Per-finding, posted by the reviewer	No (informational)
Hostile Review Thread Gate	Deterministic: unresolved hostile-reviewer threads exist	Fails until resolved (not yet a required context)
degraded verdict	Fewer than 2 models succeeded (infra)	No

Powered by omniintelligence.review_pairing.cli_review — multi-model adversarial review: qwen3-review, qwen3-review-b, glm-review (OMN-8468/OMN-8524/OMN-17492)