test(coverage): re-ratchet domain coverage floors past their 5pp drift trigger

docs/branch-review-records/023564085ea84fa8f81b78f3ec3e137053c64dbd30e5af948b0b041bd6ba537b.record.md

@@ -0,0 +1 @@
+| 2026-08-14 | PR-1964 | 1dfc0c9b984152f6518060712ab050311f3f77c9 | vitest.config.mts, docs/maturity-backlog-workorders.md (ledger #192 / maturity X6 coverage floors) | APPROVED — coverage-threshold re-ratchet, no RAG/clinical behaviour change; verify:pr-local green except pre-existing unrelated check:medication-lexicon-report staleness | test:coverage,verify:pr-local |

docs/maturity-backlog-workorders.md

@@ -158,15 +158,37 @@ structural change, not a single mixed PR.
 - **Verification:** `npm run check:drift` + the CI `db-reset-verify` replay; **live-DB work is
  confirmation-required** per the AGENTS.md provider boundary.
 
-### X6 · Raise coverage floors for clinical domains — `OPEN`
+### X6 · Raise coverage floors for clinical domains — `DONE`
 
 - **Outcome:** higher targeted thresholds where correctness matters most (clinical-safety,
  retrieval, answer) than the current global 38–50%.
-- **Approach:** add per-path coverage thresholds in `vitest.config.mts`; add the targeted tests
- needed to clear them (ratchet up, don't drop the bar retroactively).
-- **Files:** `vitest.config.mts` (+ new `tests/*` specs).
-- **Risk:** MEDIUM — needs real tests, not just a threshold bump.
-- **Verification:** `npm run test:coverage` meets the new per-path floors.
+- **Landed (ledger #192, 2026-08-14):** `vitest.config.mts` already carried four per-domain
+ coverage-threshold groups — retrieval (`clinical-search`, `retrieval-selection`,
+ `answer-ranking`, `clinical-value-binding`, `medication-entities`,
+ `rag-candidate-sources`, `rag-context-selection`, `rag-retrieval-variants`,
+ `rag-routing`), evidence/verification (`answer-verification`, `evidence`,
+ `evidence-relevance`, `rag-claim-support`, `rag-evidence-gates`,
+ `rag-quote-verification`, `rag-source-segmentation`), core RAG/answer (`rag`,
+ `rag-extractive-answer`, `rag-comparison`, `rag-answer-support`), and
+ clinical-safety/governance (`clinical-safety`, `source-governance`, `source-review`,
+ `clinical-review-queue`, `answer-response`) — plus a whole-repository floor, all
+ CI-enforced by the required `coverage` job (`npm run test:coverage`) on every PR that
+ touches executable scope. This closed out most of X6 before this ledger item was
+ worked; re-measuring on 2026-08-14 confirmed the floors were honest (no false
+ failure) and found two branch thresholds that had drifted past this file's own
+ documented 5pp re-ratchet trigger, so those two were raised (never lowered) to
+ restore headroom: evidence/verification branches 81 → 84 (measured 86.04%, gap was
+ 5.04pp) and core RAG/answer branches 72 → 76 (measured 78.22%, gap was 6.22pp). No
+ other value in this file needed a change — every other floor already sat within the
+ 5pp band. No file in the four domain groups showed a genuine untested critical path:
+ the per-file low points were 76.4% statements (`rag-quote-verification.ts`) and 57.7%
+ branches (`rag-answer-support.ts`), well clear of the "under 30%" bar this ledger item
+ treats as scary, so no padding tests were added, per the ledger's own stop condition.
+- **Files:** `vitest.config.mts`, `docs/maturity-backlog-workorders.md`.
+- **Risk:** MEDIUM — needs real tests, not just a threshold bump. (No new tests were
+ needed this pass; the floors were already backed by real, passing coverage.)
+- **Verification:** `npm run test:coverage` — 606 files / 6549 tests passed, all
+ thresholds green, both before and after the two branch-floor raises.
 
 ### X7 · Complete the `src/lib` domain-directory reorg — `OPEN`
 
@@ -291,7 +313,7 @@ collaborators join — `AGENTS.md` + the PR template already carry that load.
 | X3 Monolith decomposition | Next | IN PROGRESS (DocumentViewer #1025 + Dashboard #1034/#1042/#1047 done; rag.ts open) |
 | X4 SAST-blocking on parser | Next | **DONE** (gate + policy check) |
 | X5 ACL-migration consolidation | Next | PROVIDER-GATED (DB owner) |
-| X6 Coverage floors | Next | OPEN  |
+| X6 Coverage floors | Next | **DONE** (ledger #192) |
 | X7 `src/lib` domain reorg (rest) | Next | OPEN (follow-on to X2; needs RAG flag on answer/retrieval clusters) |
 | L1 Archive one-shot scripts | Later | IN PROGRESS (#1033 archived m13/july8; refs reconciled; backfills open) |
 | L2 Action-SHA uniformity | Later | **DONE** (#992) |

vitest.config.mts

@@ -50,27 +50,42 @@ const config = {
  // Aggregate behavioral floors ratchet the full post-fixture group rather
  // than making individual large RAG modules brittle. Each value is the
  // greater of the measured whole-group floor or the broad floor + 5pp.
+ // Ledger #192 (maturity X6, 2026-08-14): re-measured against
+ // `npm run test:coverage` on this exact file set. Group B branches (81)
+ // and Group C branches (72) had drifted >5pp below measured — the same
+ // re-ratchet trigger documented above for the whole-repo floor — so
+ // those two values were raised to sit ~2pp under the fresh measurement.
+ // Every other value here already sat within that 5pp band and is
+ // unchanged: raising a floor that already tracks its baseline closely
+ // would risk a false failure from ordinary day-to-day branch-count
+ // variance, not close a real gap.
  "src/lib/{clinical-search,retrieval-selection,answer-ranking,clinical-value-binding,medication-entities,rag/rag-candidate-sources,rag/rag-context-selection,rag/rag-retrieval-variants,rag/rag-routing}.ts":
  {
+ // Measured 2026-08-14: 90.32/81.02/92.86/93.88 (stmt/branch/func/line).
  statements: 86,
  branches: 78,
  functions: 88,
  lines: 90,
  },
  "src/lib/{answer-verification,evidence,evidence-relevance,rag/rag-claim-support,rag/rag-evidence-gates,rag/rag-quote-verification,rag/rag-source-segmentation}.ts":
  {
+ // Measured 2026-08-14: 93.89/86.04/95.24/96.44. Branches raised
+ // 81 -> 84 (gap was 5.04pp, just past the 5pp trigger).
  statements: 92,
- branches: 81,
+ branches: 84,
  functions: 94,
  lines: 94,
  },
  "src/lib/rag/{rag,rag-extractive-answer,rag-comparison,rag-answer-support}.ts": {
+ // Measured 2026-08-14: 87.55/78.22/92.57/91.84. Branches raised
+ // 72 -> 76 (gap was 6.22pp, the largest in this file).
  statements: 83,
- branches: 72,
+ branches: 76,
  functions: 90,
  lines: 88,
  },
  "src/lib/{clinical-safety,source-governance,source-review,clinical-review-queue,answer-response}.ts": {
+ // Measured 2026-08-14: 96.72/85.11/98.73/98.95.
  statements: 94,
  branches: 83,
  functions: 96,