docs(issues): close out PR E (therapy-compass) — already merged, close the rollout tracking row

[BigSimmo]

Summary

• This was requested as "PR E — therapy-compass" (the last unstarted mode in the docs/filter-contract.md rollout), following the kickoff prompt in the handover artifact. Investigation found there is no code left to write: therapy-compass's convergence onto the shared ResultFilterSheet/ResultFilterTrigger/ResultFilterFacetChips is already merged into main.
• It was actually done twice, independently, by two different concurrent Claude sessions working from the same handover artifact without visibility into each other:
  • feat(therapy-compass): converge filter sheet onto the shared contract #1885 merged first — converged the trigger/sheet and resolved the onClearAll conflict (wired to clearSearchFilters, filters only, matching docs/filter-contract.md §6) but left the Topics tag-matching AND/OR behavior "mode-owned" (unchanged).
  • feat(filters): therapy-compass adopts the filter contract (PR E) #1889 merged after — redid the same convergence and additionally fixed a real bug: searchTherapies' Topics predicate was AND-within-group (a second topic narrowed instead of widening), the same defect class already fixed for document tags. feat(filters): therapy-compass adopts the filter contract (PR E) #1889 is what's live on main now.
• Re-verified this against current main HEAD rather than trusting the merge-commit messages (per this repo's "third-party fix claims stay unverified until checked" rule): src/components/therapy-compass/screens/search-screen.tsx uses the shared components exclusively, onClearAll never touches the query, Topics is a correct OR-within-group facet, Review status and Handout are independent one-option facet groups. Lint, typecheck, and the therapy-compass/services/documents/shared-component test suites all pass. result-filter-control.tsx has exactly one dense = ... definition — no duplicate or conflicting logic survived the double implementation.
• Since services (feat(filters): services adopts the filter contract (PR C, rebased — supersedes #1866) #1878, then feat(services): adopt the filter contract with real facets (PR C) #1882 — also independently duplicated), factsheets (feat(factsheets): add per-category counts; correct the eviction claim (PR D) #1888), therapy-compass (feat(therapy-compass): converge filter sheet onto the shared contract #1885/feat(filters): therapy-compass adopts the filter contract (PR E) #1889), and documents (feat(filter-contract): converge documents onto the shared filter component #1910) are all now merged, this PR queues a done request for /issues #170 — the row tracking the entire filter-contract rollout — through docs/outstanding-issues-inbox/ (this repo's new conflict-free ledger-write mechanism; direct edits to docs/outstanding-issues.md are now rejected by check:ledger-write-discipline). Reconciliation into the canonical markdown happens after this lands via npm run issues:reconcile.
• Also adds a docs/branch-review-records/ entry documenting the post-merge verification, since no ledger row existed for feat(filters): therapy-compass adopts the filter contract (PR E) #1889.
• Flagged, not fixed here: the services merge (feat(filters): services adopts the filter contract (PR C, rebased — supersedes #1866) #1878 then feat(services): adopt the filter contract with real facets (PR C) #1882) wasn't re-audited line-by-line for what survived conflict resolution. A quick check shows src/lib/registry-records.ts (last touched by an unrelated 2026-07 commit) never got a typed ServiceRecord.facets carrier — feat(filters): services adopts the filter contract (PR C, rebased — supersedes #1866) #1878's catalogPayload-based approach is what's live instead. Both approaches may be functionally equivalent, but this wasn't verified as part of this PR. Worth a follow-up if a deeper reconciliation audit is wanted.

Verification

• npm run verify:pr-local — recognized as low-risk docs/inbox scope; summary: completed: check:runtime, check:installed-lock-parity, format:changed, sitemap:check, docs:check-index, docs:check-inventory, docs:check-scripts, docs:check-links, check:branch-review-ledger, check:outstanding-issues, check:ledger-write-discipline — failed: (none) — not reached: (none). Build skipped (no build-affecting changes detected).
• Re-ran against current main HEAD, independent of the merge-commit claims:
  • npx tsc --noEmit clean.
  • npx eslint clean across therapy-compass, result-filter-control.tsx, services, factsheets, document-search-results.tsx.
  • tests/therapy-compass-search-filters.dom.test.tsx, tests/therapy-search-facets.test.ts, tests/therapy-compass-mode-wiring.test.ts, tests/search-results-header-band.dom.test.tsx, tests/document-filter-panel.dom.test.tsx, tests/service-facets.test.ts, tests/services-catalog.test.ts, tests/decoration-on-text.contract.test.ts — 8 files, 162 tests, all passed.
  • Confirmed a single dense = facetGroups.length > 3 || totalFacetOptions > 20 definition in result-filter-control.tsx (no duplicate dense-tier logic from the two therapy PRs).
• npm run check:ledger-write-discipline passes (the gate that specifically catches an improperly-applied outstanding-issues mutation — it caught my first attempt, which hand-edited the markdown directly; this PR's committed version queues the request correctly instead).

Risk and rollout

• Risk: none — no product code changes. classifyPullRequestFiles reports clinicalRisk: false, operationalRisk: false, ragRanking: false, ui: false.
• Rollback: revert this commit; the queued inbox request simply won't be reconciled.
• Provider or production effects: None.

Notes

• No new PR was needed for the actual therapy-compass code — it's already on main via feat(therapy-compass): converge filter sheet onto the shared contract #1885/feat(filters): therapy-compass adopts the filter contract (PR E) #1889. This PR is pure bookkeeping to close the loop that "begin and complete PR E" surfaced.
• npm run issues:reconcile needs to run after this merges to actually move /issues #170 into the archive table.

---

Generated by Claude Code

[supabase]

This pull request has been ignored for the connected project sjrfecxgysukkwxsowpy because there are no changes detected in supabase directory. You can change this behaviour in Project Integrations Settings ↗︎.

---

Preview Branches by Supabase.
Learn more about Supabase Branching ↗︎.

[coderabbitai]

Warning

Review limit reached

You’ve reached a temporary PR review limit under our Fair Usage Limits Policy.

Your recent review volume is higher than typical usage, so adaptive limits are currently applied.

Next review available in:41 minutes

Your organization has reached its usage spending cap. Adjust your spending cap in the billing tab.

How can I continue?

After more reviews become available, a review can be triggered using the @coderabbitai review command as a PR comment. Alternatively, push new commits to this PR.

To avoid repeated limits, reduce automatic review volume by pausing incremental auto-reviews earlier, using label-based review opt-in, excluding WIP or generated PR titles, or requesting reviews manually when the PR is ready. If your team needs uninterrupted high-volume reviews, an organization admin can enable usage-based reviews.

How do review limits work?

CodeRabbit enforces per-developer PR review limits for each organization. Most developers receive the normal plan review availability.

For paid Pro and Pro+ PR reviews, CodeRabbit uses adaptive limits for sustained high-volume activity. When a developer's recent PR review activity reaches the 95th percentile or higher among CodeRabbit users, additional reviews become available more gradually as earlier reviews age out of the rolling window.

Please refer docs for additional details.

Review details

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Pro

Run ID: 2cc733d6-db9f-4fc7-aaba-5b6a83386ca2

📥 Commits

Reviewing files that changed from the base of the PR and between 5d67c2a and fb2e630.

📒 Files selected for processing (2)

• docs/branch-review-records/f468973451ef309665b582ddc5ccdb20e6efe16873ab26396ce2813f2dad8ab5.record.md
• docs/outstanding-issues-inbox/02879c2f-f7c7-4698-bc2e-cd5555cc2f37.json

---

_{Comment @coderabbitai help to get the list of available commands.}

[github-actions]

CI triage

CI failed on this PR. Automated classification of the 2 failed job(s):

• Static PR checks — needs investigation: inspect the failing step and uploaded diagnostics; rerun only after classifying the cause.
• PR required — needs investigation: inspect the failing step and uploaded diagnostics; rerun only after classifying the cause.

Compared with main CI run #10571 (failure).

Classification is evidence routing, not permission to ignore a failure. Exact quarantined Playwright identities remain governed by the flake ledger.

[BigSimmo]

Codex review-and-fix summary

This comment supersedes the stale PR title and body.

• Final reviewed head:7c552ccad3c415b9eaa7be27bcfb0c07be45f0fb
• Latest base:794a4786baf74ced44d3d455334ba8639662bf72
• Sync and merge-tree: One permitted, non-force merge of main was pushed at 7c552cc... against base 984f42c.... No rebase or force-push occurred. main then independently advanced once through PR chore(issues): reconcile 19 inbox requests into the canonical ledger #1931, leaving the branch ahead by 10 and behind by 1. No second sync was performed. The latest-base synthetic tree dc88997657b0dd0783317afb7d42436be925d61e cleanly overlays the two unique reviewed blobs onto base tree 2be7697b48407b2cae50314a856562b59e53715e; neither final path exists on the latest base.
• Issues fixed:
  • Removed the duplicate pending done mutation for issue #170, which would make issues:reconcile reject the batch.
  • Corrected therapy chronology: feat(filters): therapy-compass adopts the filter contract (PR E) #1889 introduced the live tree; feat(therapy-compass): converge filter sheet onto the shared contract #1885 merged later with an identical tree.
  • Corrected services chronology: feat(filters): services adopts the filter contract (PR C, rebased — supersedes #1866) #1878 introduced the live tree; feat(services): adopt the filter contract with real facets (PR C) #1882 merged later with an identical tree.
  • Replaced the inaccurate content-addressed review record.
  • Replaced the speculative services merge-audit request with a focused P2 task for offline registry facet round-trip coverage.
• Dispositioned or out of scope: No therapy-compass product-code defect remained. Registry round-trip tests remain intentionally tracked as P2; no provider-backed or live-service sampling was run. PR metadata remains stale because the connector safety layer blocked title/body edits before GitHub received them.
• Independent adversarial review: A separate manual fresh-context pass was performed because CodeRabbit was unavailable due its rate/spend limit. It found the duplicate mutation and chronology defects above. No actionable finding remains in the final two-file diff.
• Threads: No review threads, submitted reviews, or remaining actionable comments.
• Decisive verification: The review-record filename matches the row's SHA-256; the record has six columns and a full commit SHA. The inbox request has valid JSON, UUID, date, action, and payload structure. GitHub commit-tree comparisons established the feat(therapy-compass): converge filter sheet onto the shared contract #1885/feat(filters): therapy-compass adopts the filter contract (PR E) #1889 and feat(filters): services adopts the filter contract (PR C, rebased — supersedes #1866) #1878/feat(services): adopt the filter contract with real facets (PR C) #1882 histories. Current-main source review covered therapy filter wiring, service facet parsing, registry mapping, ranking, and focused fixture coverage. Exact-head static CI passed formatting, runtime and installed-lock alignment, documentation inventory and links, branch-review ledger integrity, outstanding-issues integrity, and ledger write discipline.
• Exact-head required CI:PR required green; Static PR checks green; PR policy green; PR mergeability green while correctly reporting the later one-commit base advance. Production UI was skipped as expected for this documentation-only diff. No required check is pending, missing, cancelled, or failed on the reviewed head.
• Advisory checks: Secret Scan green; SAST green. CodeRabbit was unavailable due rate limiting. Supabase was non-applicable to this diff.
• Auto-merge: Remains armed with squash mode and was not changed.
• Limitations: The local runtime had no repository checkout, GitHub network access, or gh, so authoritative GitHub connector data and exact-head Actions were used. Continued hourly monitoring is unavailable after this active run ends.
• Residual risk: GitHub currently labels the PR behind because main advanced after the single authorised sync. The latest-base tree is clean, but branch protection or auto-merge may still require a user-managed update to latest-base parity. The P2 registry round-trip tests remain future work.

The PR was not merged. Final merge is left to the user.