Ten original synthetic model-output pairs with pinned prompt/evaluation bindings, reproduction tests and manifest-versioned v1.1 release. Preserve all55original controls and Scenarios001/002. Root AI review; no inference or new human approval claimed.
19 days ago
Add Failure Scenario 002: citation resolution versus stale evidence
Ten original synthetic RAG evidence fixtures with deterministic reproduction tests. Root AI reviewed. Preserve all original control records and Failure Scenario 001. No model run, real client incident or answer release.
26 days ago
Correct scenario navigation to the verified Static Space host
Add ten original synthetic authority-mismatch cases, schema, evaluator and reproduction checks. Preserve all 55 control records and existing data bytes. Root AI review; no new human review, model benchmark or real action execution claimed. Authorized by the current named publication request.
Record owner review and approve public v0.1 candidate
Functional private candidate: security 24, AIBOM 8 original proposals, drift 11 and governance 12. Preserve native source snapshots and original release schema; expose unreviewed candidate states explicitly. New mappings remain proposals pending canonical adoption and human review.
Reconcile v0.1 candidate with four published source catalogs
Pin exact GitHub source revisions and hashes; document 52 native source records and field-level compatibility gaps. Preserve zero released rows and private source-gated status. User authorized publication, but the common control schema and combined Dataset review remain unresolved.
