docs(planning): record W5 hygiene wave + STATE tick

Eight commits closed all seven items from `.planning/v1422-backlog.md`.
STATE entry documents the per-item changes; new
`phase-W5-v1423-report.md` captures the H5 deploy decision, the H7
admin view path + the first useful question it answers, and the
OpenAPI registry additions (3 routes — H4 GET + PUT, H7 POST).

Test deltas: 2191 → 2223 unit (+32), +2 integration files, no
regressions.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

.planning/STATE.md

@@ -1,7 +1,7 @@
 # v1.4.23 marathon — state log
 
-Status: Wave 4 shipped (OpenAPI gate + Coach schema + native auth + Coolify)
-Last update: 2026-05-11T01:15+02:00
+Status: Wave 5 shipped (7 hygiene items + 8 atomic commits)
+Last update: 2026-05-11T02:00+02:00
 
 > Previous milestone: v1.4.22 live (image digest
 > `sha256:865154614303…`, `/api/version=1.4.22`).
@@ -148,13 +148,49 @@ Last update: 2026-05-11T01:15+02:00
 
 ## Wave 5 — Hygiene (H1-H7)
 
-- [ ] H1 sentinel parser malformed-enum hardening
-- [ ] H2 analytics-route unbounded `findMany` pagination
-- [ ] H3 `<CoachDrawer key={prefill}>` controlled-prop refactor
-- [ ] H4 per-user prompt-tuning surface (Coach settings cog return)
-- [ ] H5 schema drift on `medication_schedules.days_of_week`
-- [ ] H6 Pearson p-value normal-approx replacement
-- [ ] H7 Coach helpful/unhelpful first-week observation view
+- [x] H1 sentinel parser malformed-enum hardening
+      (`SentinelParseResult.malformedEntries[]` + per-line typed
+      reasons; chat route splits `coach.keyvalues.parse_partial` from
+      the full-block `coach.keyvalues.parse_failed`)
+- [x] H2 analytics-route unbounded `findMany` pagination
+      (cursor-paged 5 000-row chunks via `fetchBpSeriesChunked`,
+      `analytics.bp_in_target.row_count` wide-event meta for slow-
+      query attribution; integration test seeds 6 000 rows across a
+      chunk boundary)
+- [x] H3 `<CoachDrawer key={prefill}>` controlled-prop refactor
+      (`useResettableValue` hook + pure `nextResettableValue` helper;
+      `key={prefill}` removed from the parent mount)
+- [x] H4 per-user prompt-tuning surface (Coach settings cog return)
+      (`User.coachPrefsJson` migration `0038_coach_prefs`; new
+      `GET/PUT /api/auth/me/coach-prefs`; settings cog opens a
+      right-edge `<Sheet>`; system-prompt prepends a per-user
+      OVERRIDE; snapshot reads prefs BEFORE measurement queries so
+      `excludeMetrics` filters before the snapshot lands)
+- [x] H5 schema drift on `medication_schedules.days_of_week`
+      (decision: deploy — column is referenced by 9 source files
+      across 4 user-visible surfaces; migration
+      `0039_medication_schedule_days_of_week`, NULL = daily)
+- [x] H6 Pearson p-value normal-approx replacement
+      (raised `MIN_PAIRED_N` from 14 → 20 — conservative patch;
+      rigorous incomplete-beta queued as a v1.4.24 candidate)
+- [x] H7 Coach helpful/unhelpful first-week observation view
+      (polymorphic `RecommendationFeedback.target_type` migration
+      `0040_recommendation_feedback_target_type`; new POST
+      `/api/insights/chat/messages/:id/feedback`; aggregator slice
+      `buildCoachFeedbackBuckets` by (promptVersion, tone, verbosity);
+      admin section `/admin/coach-feedback` with the H7 i18n bundle
+      both EN + DE)
+
+W5 commits (8): 58ae9bc, fa07748, 9413d29, 3f60c81, 0eda1de, 1faee95,
+05c7f14, plus this STATE tick.
+
+Test deltas: 2191 → 2223 unit (+32). Integration suite added 3
+files: `coach-prefs`, `coach-feedback`, plus the 6000-row chunk
+boundary regression in the existing `bp-in-target` test.
+
+OpenAPI registry coverage: +3 routes (GET / PUT
+`/api/auth/me/coach-prefs`, POST
+`/api/insights/chat/messages/{id}/feedback`).
 
 ## Wave 6 — Multi-agent QA + Product-Lead review
 

.planning/phase-W5-v1423-report.md

@@ -0,0 +1,90 @@
+# v1.4.23 Wave 5 — hygiene (H1-H7) + STATE tick
+
+Closes the seven hygiene items deferred from
+`.planning/v1422-backlog.md`. Eight atomic commits; develop branch
+clean, all unit + drift gates green.
+
+## Commits
+
+1. `58ae9bc` — `feat(coach): partial-malformed observability on
+   sentinel parser` (H1)
+2. `fa07748` — `perf(api): chunked BP aggregate replaces unbounded
+   findMany` (H2)
+3. `9413d29` — `refactor(coach): drawer prefill becomes a controlled
+   prop` (H3)
+4. `3f60c81` — `feat(coach): per-user prompt-tuning surface (settings
+   cog returns)` (H4 — prisma + route + UI)
+5. `0eda1de` — `chore(schema): deploy medication_schedules.days_of_week
+   column` (H5)
+6. `1faee95` — `fix(insights): tighter Pearson surfacing gate (n>=20)`
+   (H6)
+7. `05c7f14` — `feat(coach): per-message thumbs feedback + admin
+   aggregate view` (H7 — schema + route + UI)
+8. (this commit) — `docs(planning): record W5 hygiene wave + STATE
+   tick`
+
+## H5 decision
+
+Deploy the column. `MedicationSchedule.daysOfWeek` is referenced by
+nine source files across the form, card, intake list, reminder
+dispatcher, and gamification path. Removing it would force a UI
+rewrite of the recurrence picker and break existing test fixtures.
+Migration `0039_medication_schedule_days_of_week` adds the column
+nullable, default NULL. NULL means "daily" — `parseScheduleRecurrence(null)`
+already returns the all-7-days set so existing rows get the correct
+semantics without backfill.
+
+## H7 admin view
+
+Path: `/admin/coach-feedback`. Renders the Coach-only buckets sliced
+by (promptVersion, tone, verbosity) with helpful / not-helpful / n /
+helpful-rate columns. First useful question it answers in the
+v1.4.23 first-week window: "Is the v1.4.22 prose rewrite landing
+well, or did the warm tone overshoot?" If the helpful-rate for
+PROMPT_VERSION 4.23.x drops below 50% within the first 100 ratings,
+v1.4.24 walks the persona back.
+
+## OpenAPI registry — additions
+
+Three new routes registered:
+
+- `GET /api/auth/me/coach-prefs` (H4)
+- `PUT /api/auth/me/coach-prefs` (H4)
+- `POST /api/insights/chat/messages/{id}/feedback` (H7)
+
+Spec regenerates without drift; `pnpm openapi:check` passes.
+
+## Test deltas
+
+- Unit: 2191 → 2223 (+32)
+- Integration: +2 new files (`coach-prefs.test.ts`,
+  `coach-feedback.test.ts`) plus a 6 000-row chunk-boundary
+  regression added to `bp-in-target.test.ts`
+
+## Verification gates
+
+- `pnpm typecheck` — clean across all 8 commits
+- `pnpm lint` — 0 errors, 21 warnings (all pre-existing in
+  unrelated files)
+- `pnpm test --run` — 256 files / 2223 tests passing
+- `pnpm openapi:check` — spec in sync
+
+## Carry-overs
+
+None — all seven hygiene items shipped.
+
+## Notes for v1.5
+
+- The Pearson surfacing-gate raise (n≥14 → n≥20) means some
+  borderline correlation cards STOP rendering. If the v1.4.16 B5e
+  feedback aggregator shows users miss those cards, drop in the
+  rigorous incomplete-beta replacement in v1.4.24 (queued as the
+  follow-up).
+- The Coach settings cog opens a small surface today (4 controls).
+  v1.4.24+ can iterate based on the H7 admin dashboard's first-week
+  signal — the design review pushed back on per-message controls in
+  v1.4.23, defer that decision to v1.5 P3.
+- Per-user `coachPrefsJson` is read on every Coach turn — no caching
+  layer to invalidate. The settings sheet writes the
+  `["coach-prefs"]` query cache directly so the next reply honours
+  the new defaults without a page reload.