Skip to content

feat(dashboard): add fail-soft AI enrichment#11

Merged
F-e-u-e-r merged 1 commit into
mainfrom
feat/p3-dashboard-enrichment
Jun 22, 2026
Merged

feat(dashboard): add fail-soft AI enrichment#11
F-e-u-e-r merged 1 commit into
mainfrom
feat/p3-dashboard-enrichment

Conversation

@F-e-u-e-r

Copy link
Copy Markdown
Owner

Summary

  • Load ai-annotations-meta.json and ai-annotations.json after canonical data succeeds, using crypto-free shared strict contracts and browser SHA-256 verification.
  • Join annotations only by node_id; add category and AI-tag facets, canonical URL state, AI-aware lexical search, coverage count, and clearly secondary card enrichment.
  • Add ADR-001: vector semantic search is deferred; lexical AI-enriched search ships now.
  • Keep canonical loading fail-closed and AI fail-soft.

Safety

  • Invalid, missing, hash-mismatched, or schema-invalid AI artifacts render as no AI; the canonical dashboard remains usable.
  • AI cannot override canonical repository data.
  • @starred/ai-schema/contracts is browser-safe; node:crypto stays in node-only hash/build code.

Verification

  • 455 tests across 65 files passed.
  • Typecheck, ESLint, Prettier, dashboard production build, and AI schema drift checks passed.

Scope

Delivers P3.4 implementation and P3.5 semantic-search ADR. P3 live closeout remains pending: add verify-ai-provenance to the main required checks, run one real bounded executor artifact PR, verify Pages, then prove a no-churn replay.

Load and validate optional AI annotations with browser-safe shared contracts, then join them to canonical repositories by node ID.

Add AI category/tag facets, URL state, enriched lexical search, secondary repository-card summaries, and a semantic-search ADR. Preserve canonical fail-closed loading while keeping AI artifacts fail-soft; live executor closeout remains operationally pending.
@F-e-u-e-r F-e-u-e-r merged commit 67076de into main Jun 22, 2026
4 checks passed
F-e-u-e-r added a commit that referenced this pull request Jun 22, 2026
  ## Summary
  - Update workspace documentation for the implemented P3 AI contracts, trusted planner, provenance checks, state workflow, and fail-soft dashboard enrichment.
  - Document optional AI artifacts and clarify that model-calling executors run outside CI.
  - Record that CI, `verify-agent-artifacts`, and `verify-ai-provenance` are required on `main`; retain the real executor run, Pages verification, and no-churn
  replay as pending operational closeout.

  ## Verification
  - `prettier --check README.md docs/P3-ai-spec.md` passed.
  - Documentation-only follow-up to merged P3.4/P3.5 PR #11.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant