feat(hub): human proposal evaluation, triggers, evaluator role, canister V3
- Add evaluation flow: rubric checklist, outcome (pass/fail/needs_changes), optional grade, comments, gate via HUB_PROPOSAL_EVALUATION_REQUIRED or data/hub_proposal_policy.json; approve waiver + audit entries. - Deterministic review triggers (hub_proposal_review_triggers.json): phrases, path prefixes, labels; review_queue/review_severity; proposal_auto_flagged audit. - Add evaluator role; HUB_EVALUATOR_MAY_APPROVE for approve; optional async LLM review hints (KNOWTATION_HUB_PROPOSAL_REVIEW_HINTS). - Gateway: merge policy/triggers on POST proposals; settings parity; hosted hints. - ICP: StableStorage V3 migration, review fields, POST .../review-hints. - Hub UI: evaluation form, filters, settings/how-to copy; MCP hub proposal tools. - Docs: PROPOSAL-LIFECYCLE, HUB-API, OpenAPI, TEAMS; tests for evaluation and triggers.
0 comments
muse hub commit comment sha256:7b212ddfb411f4a1ab694efd64f98f70ca6794524d8ec9f8be2bc4023d85b107 --body "your comment"
No comments yet. Be the first to start the discussion.