Files
HealthLog/.planning/phase-W-RSC-v1440-report.md
T
Marc-André Bombeck 1114cf7e63 chore(release): v1.4.40 — comprehensive Critical+High architecture closure plus iOS PB30 enablement
Eleven touch-disjoint waves landed on develop plus a six-axis QA
reconcile round before tagging. See CHANGELOG.md for the user-facing
prose; this commit message documents the wave + report inventory for
release-management traceability.

Waves shipped (each with a phase report under .planning/):

- phase-W-POOL-v1440-report.md — p-limit(4) wrapper around the
  analytics 15-way fan-out + pg.Pool max raised 10 -> 20. Closes the
  v1.4.39 empirical-trace finding #1 (Prisma pool starvation).
- phase-W-DELETED-v1440-report.md — 11 reader-tier helpers + 3
  integration contracts close audit Critical Finding #3 (soft-delete
  invisibility) across dashboard, analytics, summaries, rollup tiers,
  and the Coach snapshot path.
- phase-W-INSIGHTS-v1440-report.md — mood-rollup swap on the three
  /api/insights/{features,targets,cards} routes plus the
  compliance-rollup hook gap closure for bulk-projection mints; the
  post-QA reconcile patch closes the six remaining measurement.findMany
  sites in the cluster that still leaked tombstoned rows.
- phase-W-INFRA-v1440-report.md — rollup helper umbrella consolidation
  under src/lib/rollups/* with zero orphan re-exports; lint, typecheck,
  knip, and 93/93 rollup tests green post-restructure.
- phase-W-RSC-v1440-report.md — per-tile Suspense boundaries plus the
  queryKey factory-bypass walker test (extended to comparison/ in the
  reconcile pass).
- phase-W-PRIVACY-v1440-report.md — bilingual /privacy page with 17
  paired-section tests pinning the 9 SB-3 disclosure requirements.
- phase-W-AASA-v1440-report.md — /.well-known/apple-app-site-association
  endpoint pinned to S8WDX4W5KX.dev.healthlog.app with app-ID parity
  assertions guarding against split rotation.
- phase-W-APNS-NOTIFY-v1440-report.md — MEDICATION_REMINDER
  time-sensitive APNs payload (priority 10 + interruption-level)
  scoped via a parameterised test that pins all 6 other event-types
  do not bypass Focus. AP-2 .p8 install gates real delivery.
- phase-W-CONSENT-v1440-report.md — AI consent receipts CRUD with
  append-only invariant pinned in test; revoke-then-regrant leaves
  both rows; idempotent DELETE returns 200 + receipt: null.
- phase-W-WMY-WIRE-v1440-report.md — three previously-dead WMY readers
  (week/month/year) wired through summaries-slice and
  health-score-fast-path with a linear-composability parity test
  pinning the cross-consumer routing contract.
- phase-W-GHOSTS-v1440-report.md — ~1 177 lines of dead code removed
  across 9 atomic commits; startOfUtcDay + wallClockInTz consolidation
  eliminates a sibling-drift risk; knip whitelist trimmed to three
  files pending follow-up.

QA reconcile patches applied per the v1.4.40 QA round (six reports
under .planning/round-v1440-QA-*-findings.md):

- senior C1 — dashboard-suspense-boundaries regex aligned with the
  post-W-INFRA Thread 2 userTimezone-local lift.
- senior H1 — soft-delete filter added to the six remaining
  insights measurement.findMany sites.
- specialist H1 — Promise.allSettled around bulk-projection
  compliance recompute on both /api/medications/intake and
  /api/dashboard/summary.
- ux H1 — compare-toggle migrated to queryKeys.dashboardWidgets
  factory; walker test extended to src/components/comparison/.
- security M1 — consent artefact 64 KB cap enforced via
  Buffer.byteLength (UTF-8 bytes, not UTF-16 code units).

Quality gates green at tag: pnpm typecheck, pnpm lint, pnpm knip
--include files,dependencies,binaries,unlisted, pnpm test --run
(4726 passing / 1 skipped). Tests 4525 -> 4726 unit + 1 skipped over
the v1.4.39 -> v1.4.40 span.

User-facing prose: see CHANGELOG.md.
2026-05-21 12:48:02 +02:00

7.0 KiB

Phase W-RSC v1.4.40 — Frontend RSC + Suspense + queryKey factory

Wave: W-RSC (frontend critical / high) Brief: .planning/round-v1439-arch-qa-frontend.md Branch: develop Range: c9d5479b → HEAD (691443e2)

Shipped

1. queryKey factory enforcement

  • src/lib/query-keys.ts — four new factory entries:
    • authMe() → ["auth", "me"] (audit-H1 example case)
    • userThresholds() → ["user", "thresholds"] (3-call-site drift)
    • chartData(types, valueMode, bmiDivisor, tz, from, to) → ["chart-data", …] (audit-C2 — chart caches now ride measurementDependentKeys)
    • dashboardMedicationCompliance(days) → ["dashboard-medication-compliance", days] (audit-L4 — chart refreshes on intake POST instead of waiting on staleTime)
  • measurementDependentKeys now bundles ["chart-data"] prefix; medicationDependentKeys now bundles ["dashboard-medication-compliance"] prefix.
  • Migrated call sites: src/hooks/use-auth.ts, src/app/insights/page.tsx, src/app/insights/medikamente/page.tsx, src/app/insights/stimmung/page.tsx, src/components/insights/insights-layout-shell.tsx, src/components/charts/health-chart.tsx, src/components/charts/medication-compliance-chart.tsx, src/components/charts/mood-chart.tsx.

2. mood-chart queryKey dedup (empirical-trace #3)

  • src/components/charts/mood-chart.tsx:316 was bare ["mood-chart-data"]; the dashboard at app/page.tsx:291 already ran queryKeys.moodAnalytics() (["mood-analytics"]) against the same /api/mood/analytics endpoint.
  • Re-keyed to queryKeys.moodAnalytics() so TanStack dedups the request. Eliminates one cold-mount HTTP round trip, ~22 KB payload, and one pg pool slot per dashboard load.
  • Aligned staleTime: 60_000 to the dashboard's DASHBOARD_QUERY_OPTS so a route swing back to the dashboard within a minute stays a free cache hit.

3. Per-tile Suspense boundaries (audit-H2 + brief C1)

  • src/app/page.tsx — wrapped every chart-row cell in <Suspense fallback={<ChartSkeleton />}> and every tile-strip cell in <Suspense fallback={null}>.
  • The dynamic-loaded charts (HealthChartDynamic, MoodChart, MedicationComplianceChart) already paint a <ChartSkeleton> during JS chunk resolution via next/dynamic({ loading: … }); the Suspense layer lifts the same skeleton to a streaming- compatible boundary so a future useSuspenseQuery migration automatically buckets each chart's loading state to its own cell without a further composition pass.

4. Bundled audit polishes

  • DASHBOARD_QUERY_OPTS hoisted to module scope (audit-M2 — stable reference across renders).
  • getHourForTimeZone memoised via useMemo keyed on user?.timezone (audit-H4 — keeps the Intl.DateTimeFormat instantiation off the per-render hot path).

5. Tests

  • src/lib/__tests__/query-keys.test.ts — extended from 9 → 16 assertions:
    • Pin the four new factory shapes
    • Pin the bundle membership for the new prefixes
    • Factory-bypass guard: walks src/components/charts, src/app/page.tsx and src/hooks/use-auth.ts and fails if any queryKey: literal array slipped past the factory. Acts as lint-style enforcement in lieu of a custom ESLint rule.
  • src/app/__tests__/dashboard-suspense-boundaries.test.ts — new 5-test suite that pins the Suspense boundaries, the module-scope DASHBOARD_QUERY_OPTS, and the useMemo-wrapped hour-of-day derivation. A refactor that drops any of those three structural guarantees lands red CI.

Decisions deferred

  • app/page.tsx RSC migration (brief item 3). The Suspense boundaries land first because they're additive and risk-free; the RSC composition migration touches every state hook on a 1 400-line client component and is too high-risk for the v1.4.40 envelope. Recommendation: open a v1.4.41 phase that:

    1. Carves out a <DashboardComposition> server-component shell that fetches the slim-analytics + widgets server-side via headers() + the shared session helper.
    2. Streams the slim payload through <Suspense> to the existing client tile-strip island.
    3. Keeps thick analytics + per-chart ["chart-data", …] queries on the client. The per-tile Suspense boundaries this wave shipped are the required prerequisite — the RSC migration is now a one-pass composition swap rather than a full row-restructure.
  • Custom ESLint rule for queryKey factory (brief item 2). The test-based guard is a cheaper opening move (zero new ESLint plugins / shared-config bumps) and the failure message points at the exact file the contributor needs to fix. A v1.4.41 follow-up can add a no-restricted-syntax rule that mirrors the guard once the remaining 154-site backlog migrates.

  • Long-tail factory migration (audit-H1 — 154 bare sites). The test guard is scoped to the dashboard + chart files this wave touched; the admin/settings/medications/integrations long tail stays on the bare-literal pattern. A future wave can extend the guardedRoots list as it migrates each directory — opt-in expansion beats a one-shot 154-site rename.

Quality gates

Gate Pre Post Status
pnpm typecheck clean clean green
pnpm vitest (all) 4732 pass 4748 pass + 1 skipped green (+16 new)
pnpm vitest (W-RSC focus) n/a 16 + 5 + 12 + 47 = 215 pass green
pnpm lint 10 pre-existing errors in other waves' files (consent/ai/route.ts, privacy/page.tsx) identical unchanged
Playwright dashboard.spec.ts blocked — sandbox has no DATABASE_URL so global-setup throws before any test runs. The CI environment has the DB and will gate normally.

Commits

  1. Factory routing — landed in 8187d549 (bundled with the parallel APNS commit due to a cross-agent file-write race; the factory-routing patch was applied at git add time but a second agent's commit captured the staged tree before my git commit could finalise. Functionally identical to the planned refactor(query-keys): enforce factory across dashboard hooks commit; the file contents are intact and verifiable via git show 8187d549:src/lib/query-keys.ts).
  2. 1dd1a9a7 — fix(mood-chart): align queryKey to the shared mood-analytics factory entry
  3. 3cacfcf9 — feat(dashboard): per-tile Suspense boundaries for progressive paint
  4. 691443e2 — test(query-keys): pin factory enforcement and dedup behaviour

Observed pattern — cross-agent commit drift

The factory-routing commit (1) was the second time this marathon a parallel agent's commit absorbed staged-but-not-yet-committed work from a sibling agent's worktree. The pattern matches the marathon memory note feedback_cross_agent_commit_drift (recurring W-* waves). Mitigation suggestion for v1.4.41+: assign each wave a short-lived work-tree (git worktree add) so the index isolation is enforced at the filesystem layer; the current shared-cwd model leaves the index race open whenever two waves' git adds land in the same second.