chore: remove REQUIREMENTS.md for v1.1 milestone

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
chore: archive v1.1 milestone files
2026-06-18 22:04:01 -04:00 · 2026-06-18 22:03:56 -04:00 · 2026-06-18 21:32:06 -04:00 · 2026-06-18 21:21:26 -04:00 · 2026-06-18 21:19:45 -04:00 · 2026-06-18 21:16:52 -04:00
27 changed files with 1443 additions and 587 deletions
@@ -46,12 +46,12 @@ jobs:
      - name: Enable pnpm
        run: corepack enable pnpm
-      # actions/cache@v4 is intentionally omitted — probe (D-PROBE-04) showed it
+      # actions/cache@v4 intentionally omitted (D-PROBE-04). Installs now target the
-      # times out on this runner (socket hang-up between runner container and job
+      # host-mounted pnpm store at /pnpm-store (see act_runner config.yaml container.options).
-      # container cache server). pnpm install without cache takes ~30s; acceptable.
+      # Without the host mount the flag still works — pnpm creates an ephemeral store there.
      - name: Install dependencies
-        run: pnpm install --frozen-lockfile
+        run: pnpm install --frozen-lockfile --store-dir /pnpm-store --prefer-offline
      - name: Lint
        run: pnpm lint
@@ -104,10 +104,11 @@ jobs:
      - name: Enable pnpm
        run: corepack enable pnpm
-      # actions/cache@v4 intentionally omitted — same reasoning as fast-checks job (D-PROBE-04).
+      # actions/cache@v4 intentionally omitted (D-PROBE-04). Installs now target the
      # host-mounted pnpm store at /pnpm-store (see act_runner config.yaml container.options).
      - name: Install dependencies
-        run: pnpm install --frozen-lockfile
+        run: pnpm install --frozen-lockfile --store-dir /pnpm-store --prefer-offline
      # Pitfall 11: service container healthy != MariaDB accepting connections.
      # No mysql CLI in the runner image (D-PROBE-03); poll via the already-installed
@@ -181,6 +182,9 @@ jobs:
      DB_USER: familysync
      DB_PASSWORD: testpass
      DB_NAME: familysync
      # Persist Playwright browser binaries across runs via host-mounted /ms-playwright.
      # Without the host mount CI still works — binaries are downloaded to the ephemeral dir.
      PLAYWRIGHT_BROWSERS_PATH: /ms-playwright
    steps:
      - uses: actions/checkout@v4
@@ -191,10 +195,11 @@ jobs:
      - name: Enable pnpm
        run: corepack enable pnpm
-      # actions/cache@v4 intentionally omitted — same reasoning as fast-checks job (D-PROBE-04).
+      # actions/cache@v4 intentionally omitted (D-PROBE-04). Installs now target the
      # host-mounted pnpm store at /pnpm-store (see act_runner config.yaml container.options).
      - name: Install dependencies
-        run: pnpm install --frozen-lockfile
+        run: pnpm install --frozen-lockfile --store-dir /pnpm-store --prefer-offline
      # Pitfall 11: service container healthy != MariaDB accepting connections.
      # No mysql CLI in the runner image (D-PROBE-03); poll via the mysql2 driver
@@ -270,7 +275,9 @@ jobs:
      # Install Playwright browsers with system deps BEFORE starting the API, so the long
      # browser download does not run during the API's lifetime.
      # Must run from apps/pwa/ where @playwright/test is installed (D-PROBE-05 confirmed exit 0).
-      # Do NOT cache browser binaries — Playwright explicitly recommends against it in CI.
+      # PLAYWRIGHT_BROWSERS_PATH=/ms-playwright (job-level env) persists binaries across runs via
      # the host-mounted dir. The --with-deps apt step cannot be cached; baking a runner image
      # with browsers preinstalled would also drop the --with-deps apt step (future optimization).
      - name: Install Playwright browsers
        run: npx playwright install --with-deps webkit chromium
        working-directory: apps/pwa
@@ -462,7 +469,8 @@ jobs:
            --exit-code 1
      # ── pnpm audit + outdated (code-change PRs only, D-12) ───────────────────
-      # actions/cache@v4 intentionally omitted — same reasoning as fast-checks job (D-PROBE-04).
+      # actions/cache@v4 intentionally omitted (D-PROBE-04). Installs now target the
      # host-mounted pnpm store at /pnpm-store (see act_runner config.yaml container.options).
      - uses: actions/setup-node@v4
        if: needs.changes.outputs.code == 'true'
@@ -475,7 +483,7 @@ jobs:
      - name: Install dependencies
        if: needs.changes.outputs.code == 'true'
-        run: pnpm install --frozen-lockfile
+        run: pnpm install --frozen-lockfile --store-dir /pnpm-store --prefer-offline
      - name: Dependency audit (blocking on High+Critical)
        if: needs.changes.outputs.code == 'true'
@@ -88,6 +88,11 @@ jobs:
      # Build from REPO ROOT (T-08-10): the Dockerfile copies the pnpm workspace manifest +
      # lockfile from the root context; building from apps/api/ would fail to find them.
      - name: Build production image
        # DOCKER_BUILDKIT=1 is required: the Dockerfile uses `RUN --mount=type=cache`
        # (BuildKit) to persist the pnpm store across builds. The legacy builder would
        # fail on that syntax. BuildKit is default on Docker 23+, set explicitly for safety.
        env:
          DOCKER_BUILDKIT: '1'
        run: |
          set -euo pipefail
          docker build --target production \
@@ -1,5 +1,34 @@
 # Milestones
 ## v1.1 Operability & Polish (Shipped: 2026-06-18)
 **Scope:** 14 phases (7–20), 57 plans, ~110 tasks. Continues v1.0 numbering; merged to `main` across a series of phase PRs (latest #27).
 **Delivered:** Turned the v1.0 MVP into a configurable, administrable, and maintainable app — guided first-run setup, in-app role-gated admin, per-event reminders, near-instant write-back, local-auth (no-OIDC) mode, auto timezone — backed by a full self-hosted Gitea CI/CD pipeline (mobile + desktop e2e, security scanning, image hygiene, Docker publish) and a real lint gate. No more hand-editing env files or the database.
 **Key accomplishments:**
 - **Phase 7 — Mobile Test Harness:** Playwright harness (`@playwright/test`) with an iPhone/WebKit + Pixel/Chromium device matrix, SW-block, env-driven baseURL, deterministic dev-DB seed, and `DEV_AUTH_BYPASS` auth; layout/calendar/lists specs assert tap-targets, overflow, and populated/empty/error states. TEST-01/02. (Consumed by Phase 8 CI.)
 - **Phase 8 — Gitea CI:** Self-hosted Gitea Actions pipeline — parallel `fast-checks` (lint/typecheck/PWA unit) + `api` (MariaDB 11 service container + migrate + DB-backed tests) + `harness` (dev-stack bring-up + Phase 7 specs on both profiles) gating every PR to `main`, plus a publish job pushing the API production image (`:latest` + `:v1.1-<sha>`, `--password-stdin`). CI-01/02.
 - **Phase 9 — Faster Write-Back:** Event-driven outbox drain via a zero-dependency in-process EventEmitter (`outboxTrigger.ts`) — committed enqueues fire `signalOutboxDrain()` so edits land in ~1–2s instead of ~15s, preserving optimistic-202, create-before-delete, per-uid exactly-once, and the 15s fallback sweep. CAL-15.
 - **Phase 10 — Admin Role & Settings:** v1.1 DB foundation (`users.is_admin`, `member_credentials.provider_type`, `calendar_events.reminder_lead_minutes`, `app_config`); DB-backed `requireAdmin` gating all `/api/admin/*`; one shared `validateEncryptAndStoreCredential` (CalDAV PROPFIND + AES-256-GCM) for admin rotation + member self-service; gated `/admin` PWA route. ADMIN-01/02/03.
 - **Phase 11 — Per-Event Reminders:** Per-event reminder picker (None / 5m … 2d, all-day → 9 AM local) serialized as a VALARM, with a variable-lead scheduler (`uid:dtstartMs` dedup, dropped the hardcoded 15-min/shared-only restriction) that honors each event's lead, fires nothing without an alarm, and preserves VALARMs set in other clients. CAL-13/14, NOTIF-04/05/06.
 - **Phase 12 — Initial Setup Wizard:** Pre-auth `/api/setup/*` first-run wizard validating DB / VAPID / OIDC / app-password before completion, generating env secrets (never persisted to DB), promoting the completing user to admin, and locking with a 423 guard on every invocation. SETUP-01/02/03/04.
 - **Phase 13 — Real Lint Gate (ESLint):** ESLint flat config (typescript-eslint + React) across both apps + a Prettier `format:check` gate, turning the hollow `--if-present` no-op into a CI lint gate that actually fails; full first-run baseline cleanup to green.
 - **Phase 14 — Desktop E2E Coverage:** Added a `desktop` (Desktop Chrome, no-touch) Playwright project and made the mobile-authored specs desktop-safe, so the CI regression gate validates desktop as well as iphone/pixel.
 - **Phase 15 — Doc-Only CI Skip + Markdown Lint:** `dorny/paths-filter` classifies each PR so doc-only changes skip the slow `api`/`harness` jobs, with an always-running `gate` aggregate (avoids the required-check deadlock) and markdownlint-cli2 added to `fast-checks`.
 - **Phase 16 — CI Dependency Audit, Security & Image Hygiene:** Boot-time refuse-to-boot guard + baked `NODE_ENV=production` confining `DEV_AUTH_BYPASS` to dev; `pnpm audit` gate with GHSA waiver allowlist + tiered outdated report; eslint-plugin-security; gitleaks (clean 613-commit baseline) + `.dockerignore`; publish-time image-hygiene assertions. SEC/DEP/IMG/CI-03.
 - **Phase 17 — UI Optimization & Polish:** Fixed the long-standing phone BottomTabBar/FAB overlap (with a CI regression guard), shipped the real FamilySync logo + full favicon/PWA-icon set + brand accent, restructured `tokens.css` into a themeable semantic-token layer (light-only groundwork), and added a logout control + desktop-centered sheets + admin toasts.
 - **Phase 18 — Auto Timezone Detection:** Made the household timezone an explicit, stored, browser-auto-detected, admin-changeable setting (`getHouseholdTimezone(db)` + IANA validation), routing the all-day "9 AM local" reminder computation through it instead of the implicit `process.env.TZ`.
 - **Phase 19 — Local Auth (No-OIDC Mode):** Full local username/password account model (scrypt + stateless `local-session` JWT cookie, `local_credentials` table, rate-limit/lockout, login/logout, admin create/reset, self-change, OIDC-link, break-glass CLI) coexisting with the Authelia OIDC path — removing the hard dependency on a deployed Authelia.
 - **Phase 20 — Admin Member Editor & Form Declutter:** Replaced per-row Rotate/Reset buttons with a single tappable member-editor sheet (display name + local password + app password) over a new `PATCH /api/admin/members/:id` with a last-admin guard, and collapsed the Add-member form — retiring the confusing "Rotate" copy.
 **Requirements:** 17/17 v1.1 requirements complete (TEST, CI, CAL, ADMIN, NOTIF, SETUP). Phases 13–20 were driven by decision contracts (D-IDs / AUTH-LOCAL-*) rather than REQ-IDs. Deferred to backlog: self-service onboarding (999.5), provider abstraction (999.1), multiple reminders per event (v1.2), dark mode / theming (999.20), modern styling refresh (999.21).
 **Known deferred items at close:** none carried — all phase verifications (incl. Phase 11 & 17 human-needed checks) confirmed by the operator at close.
 ---
 ## v1.0 MVP (Shipped: 2026-06-10)
 **Scope:** 6 phases, 42 plans, 68 tasks. Shipped via Gitea PR #1 (`gsd/v1.0-milestone` → `main`, 375 commits).
@@ -8,20 +8,17 @@ FamilySync is a self-hosted, Dockerized family organization hub for a two-person
 The household can see and co-edit one color-coded family calendar (shared + each member's personal) and shared lists from a single low-friction PWA — cross-ecosystem, no app store, no per-member calendar credential juggling.
-## Current Milestone: v1.1 Operability & Polish
+## Current State
-**Goal:** Make FamilySync configurable, administrable, and maintainable for real multi-member use — guided setup, in-app admin, per-event reminders, faster write-back, CI/CD, and mobile test coverage — without hand-editing env files or the database.
+**Shipped: v1.1 Operability & Polish (2026-06-18)** — 14 phases (7–20), 57 plans. Full detail in [`MILESTONES.md`](MILESTONES.md) and [`milestones/v1.1-ROADMAP.md`](milestones/v1.1-ROADMAP.md).
-**Target features:**
+v1.1 turned the v1.0 MVP into a configurable, administrable, maintainable app: guided first-run setup wizard, role-gated in-app admin (credential rotation, shared-calendar designation, member editor), per-event reminders with a variable-lead scheduler, near-instant (~1–2s) event write-back, local-auth (no-OIDC) mode, and auto timezone detection — all backed by a full self-hosted Gitea CI/CD pipeline (mobile + desktop Playwright regression, a real ESLint gate, dependency/secret/security scanning, dev↔prod image hygiene, and Docker publish). No more hand-editing env files or the database.
- **Per-event reminders** — reminder selector on the event form (incl. "none"), serialized as VALARM; scheduler honors each event's lead instead of a hardcoded 15-min, and fires nothing when an event has no alarm (was backlog 999.4)
+Deferred to backlog: self-service provider onboarding (999.5), provider abstraction (999.1), dark mode / theming (999.20), and a broader modern-styling refresh (999.21 — future milestone).
 - **Admin Settings section** — role-gated UI to manage per-member Fastmail app passwords and designate the shared calendar, replacing manual DB writes (was backlog 999.10)
 - **Initial setup wizard** — first-run validated bootstrap of env vars, VAPID keypair, DB connection, and first app password (was backlog 999.11)
 - **Faster write-back** — event-driven outbox drain so edits land in ~1s instead of up to ~15s, preserving the optimistic-202 durability guarantees (was backlog 999.13)
 - **Gitea CI** — full regression (lint/typecheck/unit/API-integration against a MariaDB service container) on PR to main + build/publish Docker image (was backlog 999.14)
 - **Mobile-browser testing** ✅ **delivered (Phase 7, 2026-06-11)** — Playwright harness, two-profile mobile matrix (iPhone/WebKit + Pixel/Chromium), DEV_AUTH_BYPASS auth, deterministic dev-DB seed; 58 specs across both profiles assert layout/state. TEST-01/TEST-02 validated. Consumed by Phase 8 CI (was backlog 999.12)
-Deferred to backlog: self-service provider onboarding (999.5) and provider abstraction (999.1). Admin-managed credentials (999.10) partially cover the multi-member credential gap in the interim.
+## Next Milestone
 Not yet defined. Start with `/gsd-new-milestone` (questioning → research → requirements → roadmap). Candidate seeds in the backlog: dark mode / theming (999.20), modern visual refresh (999.21), self-service onboarding (999.5), provider abstraction (999.1), dev-user full-app exercise without a real calendar (999.19), and acting on the CI dependency report (999.18).
 ## Requirements
@@ -39,6 +36,11 @@ Deferred to backlog: self-service provider onboarding (999.5) and provider abstr
 - [x] Faster write-back so edits reach Fastmail in ~1–2s instead of ~15s (CAL-15) — **Validated in Phase 9 (faster-write-back)**: event-driven outbox drain via a zero-dependency in-process EventEmitter (`outboxTrigger.ts`); a committed enqueue publishes a fire-and-forget `signalOutboxDrain()` that funnels through the existing `isDraining`-guarded drain with a `drainRequested` trailing-re-drain, preserving optimistic-202, create-before-delete on moves, exactly-once per uid, and the 15s `setInterval` fallback. 5/5 success criteria verified; trigger-wiring tests assert SC-1/D-05/D-07.
 - [x] Per-event reminders — choose a reminder lead per event (None / 5m / 10m / 15m / 30m / 1h / 2h / 1d / 2d, all-day → day-granularity + 9 AM fire), serialized as a VALARM, with a variable-lead scheduler that honors each event's lead (CAL-13/CAL-14, NOTIF-04/05/06) — **Validated in Phase 11 (per-event-reminders)**: pure VALARM serialization/classification layer (`buildTimedValarm`/`buildAllDayValarm`/`classifyValarms`/`extractValarms`/`computeAlertInstantUtc`); variable-lead scheduler with `uid:dtstartMs` dedup, dropped fixed-15-min/shared-only restriction, all-day 9 AM-local branch; `reminderLeadMinutes` threaded end-to-end with preserve-on-no-change (D-08); allDay-aware reminder picker with edit pre-population. Gap-closure (Plan 11-05) fixed two code-review blockers — custom/other-client VALARMs are now preserved on edit via a surfaced `reminderIsCustom` signal (CAL-14 / Pitfall 1), and the all-day push body no longer reads "Starts in 0 min" — plus post-event-trigger classification, a server-side max bound, and helper-text gating. 5/5 must-haves verified; 347 API + 206 PWA tests green. **Deferred:** live Fastmail VALARM round-trip + on-device push fire (untestable in dev — no provider connected; backlog 999.19).
 - [x] Admin role + role-gated settings surface to rotate member Fastmail app passwords and designate the shared calendar (ADMIN-01/02/03) — **Validated in Phase 10 (admin-role-settings)**: v1.1 DB foundation (`users.is_admin`, `member_credentials.provider_type`+`unique(user_id)`, `calendar_events.reminder_lead_minutes`, `app_config`) via an additive generate+migrate migration; DB-backed `requireAdmin` gating all `/api/admin/*` (client `isAdmin` UX-only, server 403 the real boundary, D-03); one shared `validateEncryptAndStoreCredential` helper for admin rotation + member self-service `/api/me/credential` (400-no-echo, session-userId only); exclusive shared-calendar designation made transactional + 404-guarded (CR-01 fix); gated `/admin` PWA route + conditional nav + `SetupBanner`. 12/12 must-haves verified; admin route-guard/nav-gating green in real Chromium (e2e 5/5). Deferred follow-ups: WR-01 bootstrap-race (Phase 12 reworks the bootstrap), broker `credentialSync.ts`/`CredentialSheet.tsx` crypto re-audit under full read access.
 - [x] Initial setup wizard — first-run validated bootstrap of env/secrets/DB/OIDC/VAPID/app-password instead of hand-editing files; locks once complete (SETUP-01/02/03/04) — **Validated in Phase 12 (initial-setup-wizard)**: pre-auth `/api/setup/*` router mounted before the OIDC guard; each input validated (DB connects, VAPID decodes to 32 bytes + pairs with the public key, OIDC discovery resolves, app password reaches CalDAV); generated secrets shown for env copy and never persisted to the DB; completing user promoted to admin and a 423 guard enforced on every invocation. First-login-claims rework in `upsertUser` (no email coupling).
 - [x] Self-hosted Gitea CI/CD + automated browser test coverage (TEST-01/02, CI-01/02) — **Validated in Phases 7/8/13/14/15/16**: a mobile (iPhone/WebKit + Pixel/Chromium) **and** desktop Playwright harness reached via `DEV_AUTH_BYPASS`; a PR pipeline gating lint (real ESLint flat config) / typecheck / unit / MariaDB-backed API integration / the headless harness; doc-only PRs skip the slow jobs via an always-running `gate` aggregate; dependency audit + gitleaks + eslint-plugin-security + dev↔prod image-hygiene assertions; and a publish job pushing the API production image on merge to `main`.
 - [x] Local-auth (no-OIDC) mode (AUTH-LOCAL-*) — **Validated in Phase 19 (local-auth-no-oidc-mode)**: full local username/password account model — scrypt hashing, stateless `local-session` JWT cookie, `local_credentials` table, rate-limit/lockout login + logout, admin create/reset member, self-change password, OIDC-link to claim a local user, and a break-glass reset-admin CLI — coexisting with the Authelia OIDC path, removing the hard dependency on a deployed Authelia for solo/small self-hosters.
 - [x] Household timezone as an explicit, stored, auto-detected, admin-changeable setting (Phase 18 D-01..D-07) — **Validated in Phase 18 (auto-timezone-detection)**: `getHouseholdTimezone(db)` with IANA validation is the source of truth for the all-day "9 AM local" reminder computation (replacing the implicit `process.env.TZ`), seeded from the browser at first run and changeable from `/admin`; browser-local display/timed-write path untouched.
 - [x] PWA visual identity + phone-layout polish + admin member editor (Phase 17 D-01..D-10, Phase 20 D-01..D-07) — **Validated in Phases 17 & 20**: fixed the phone BottomTabBar/FAB overlap (with a CI regression guard), shipped the real FamilySync logo + full favicon/PWA-icon set + brand accent, restructured `tokens.css` into a themeable semantic-token layer (light-only groundwork), added a logout control + desktop-centered sheets; and replaced the per-row Rotate/Reset buttons with a single tappable member-editor sheet over `PATCH /api/admin/members/:id` (last-admin guard), retiring the confusing "Rotate" copy.
 ### Active
@@ -100,6 +102,9 @@ Deferred to backlog: self-service provider onboarding (999.5) and provider abstr
 | **D-15:** Validate the real external topology via a **local Newt connector + test subdomain** through existing Pangolin (Mode A), not an Unraid deploy. Unraid (Mode B) reserved for go-live.                                                                                                                                                                                                                                             | Authelia OIDC + SSE pass-through behaviour live in Authelia + Pangolin/Newt, not in where the origin runs — so a local Newt rig faithfully tests both, decoupling "does the topology work" from "is it in production." Newt dials outbound (no open ports). Only shared touch is an additive, reversible Authelia client.                                                                                                                                        | ✓ Validated (Gate 2 executed live in Phase 3 / D-17)                                                                                                                        |
 | **D-16 (2026-06-05, Phase 2):** No dedicated Fastmail "broker" account. The **shared-family calendar is a calendar collection created on the operator's primary Fastmail account** (`me@lucasberger.ca`) and shared out to the wife + others via Fastmail's own calendar sharing. The app's single app password enumerates it like any other collection; the `calendars.is_shared` flag (operator-set) marks which row is the shared one. | Clarified during the Wave 2 checkpoint: "broker account" was only ever the role the primary account's app password plays. id=1 ("Calendar") is the operator's **personal** calendar, not the shared one — so it must NOT be marked `is_shared`. Aggregating each _other_ member's **personal** calendar still follows the D-09 per-member app-password model (open for Phase 3 onboarding: a member may get a personal color lane, or only the shared calendar). | ✓ Resolved (2026-06-10): "FamilySync" shared calendar created on the primary account, synced as `calendars.id=10`, marked `is_shared=1`; shared lane + reminders now active |
 | **D-17 (2026-06-07, Phase 3):** Phase 1 Gate 2 (deferred per D-14) was executed live during Phase 3 against real Authelia OIDC over Pangolin/Newt (Mode A), clearing the load-bearing iOS-standalone-login risk. The full event write path (create/all-day/recurring/edit/delete/conflict) is verified end-to-end to Fastmail.                                                                                                            | Live bring-up surfaced bugs the dev-bypass build could not (newt MTU blackhole, OIDC state-cookie race, write-path timezone/identity/join/cache bugs, all-day off-by-one, color collisions). All fixed; UX gaps captured as backlog 999.3–999.9.                                                                                                                                                                                                                 | — Validated (Gate 2, `03-GATE2-RESULTS.md`). Carried: Android install (B5), SSE smoke (Phase 4 entry gate, D-14).                                                           |
 | **D-18 (2026-06-12, Phase 9):** Faster write-back uses a **zero-dependency in-process EventEmitter** drain signal, not Redis — the drain is single-process by design; Redis stays only for list SSE. | The optimistic-202 outbox is single-process; an in-process signal funnelled through the existing `isDraining` guard preserves all durability guarantees without a new external dependency. (Redis was later removed entirely — quick 260618-smr — as it was unused at runtime.) | ✓ Validated (v1.1, Phase 9, CAL-15) |
 | **D-19 (2026-06-17, Phase 19):** FamilySync ships **local username/password auth as a first-class mode coexisting with Authelia OIDC**, not OIDC-only. | The operator runs it this way; a hard dependency on a deployed Authelia is too heavy for solo/small self-hosters. A local user can be linked to an OIDC identity later (claim flow, never email-matched per D-10). | ✓ Validated (v1.1, Phase 19) |
 | **D-20 (2026-06-11, Phase 8):** CI runs on the self-hosted Gitea runner with `runs-on: ubuntu-latest` (no self-hosted label) in Docker-executor mode; MariaDB readiness uses `healthcheck.sh --connect`, never `mysqladmin ping` (removed in MariaDB 11); the secret is `REGISTRY_PAT` (the `GITEA_` prefix is silently dropped). | Established by the runner-probe-first approach (PITFALLS 11/12); these constraints are load-bearing for every CI workflow in the repo. | ✓ Validated (v1.1, Phases 8/16, CI-01/02) |
 ## Evolution
@@ -122,4 +127,4 @@ This document evolves at phase transitions and milestone boundaries.
 ---
-_Last updated: 2026-06-18 — Phase 20 (Admin Member Editor & Form Declutter) complete; single member-editor sheet (D-01..D-07) over new PATCH /api/admin/members/:id with last-admin guard; "Rotate"/"Reset password" copy retired._
+_Last updated: 2026-06-18 after v1.1 milestone — Operability & Polish shipped (Phases 7–20, 57 plans): guided setup, in-app admin + member editor, per-event reminders, faster write-back, local-auth mode, auto timezone, and full Gitea CI/CD. Next milestone undefined — start with `/gsd-new-milestone`._
@@ -46,6 +46,52 @@ _A living document updated after each milestone. Lessons feed forward into futur
 ---
 ## Milestone: v1.1 — Operability & Polish
 **Shipped:** 2026-06-18
 **Phases:** 14 (7–20) | **Plans:** 57 | **Sessions:** not tracked
 ### What Was Built
 - A self-hosted Gitea CI/CD pipeline: PR-gating lint (real ESLint flat config) / typecheck / MariaDB-backed API integration / a mobile + desktop Playwright regression harness, plus dependency-audit / gitleaks / eslint-plugin-security / dev↔prod image-hygiene gates and a Docker publish on merge.
 - In-app operability: role-gated admin (credential rotation, shared-calendar designation, member editor), a validated first-run setup wizard, per-event reminders with a variable-lead scheduler, auto timezone detection, and ~1–2s event write-back.
 - A first-class local-auth (no-OIDC) mode coexisting with Authelia OIDC, removing the hard dependency on a deployed Authelia.
 ### What Worked
 - **Backlog → phase promotion pipeline:** most of v1.1 (999.4/10/11/12/13/14/15/16) was captured as backlog during v1.0, then promoted cleanly into scoped phases — the deferred-idea capture paid off directly.
 - **Runner-probe-first for self-hosted CI (PITFALL 12):** probing `node`/`pnpm`/Docker/registry access on the Gitea runner *before* authoring any test/build steps surfaced every fork answer (Docker-executor, `ubuntu-latest`, artifact-fork, `REGISTRY_PAT` naming) up front and avoided blind CI iteration.
 - **Zero-dependency in-process solutions:** the EventEmitter outbox-drain signal (CAL-15) hit the latency goal with no new infra; the project later removed Redis entirely as unused.
 - **TDD discipline on the admin/auth chain** (Phases 10/11/12/19) kept the role boundary and credential-handling correct, with route-level 403/423/409 guards asserted in tests.
 ### What Was Inefficient
 - **Dev user can't exercise calendar features end-to-end:** `DEV_AUTH_BYPASS` user 1 has no `member_credentials`/calendars, so per-event reminders (Phase 11) could only be verified via tests + a route-mocked smoke, not hands-on by the operator (→ backlog 999.19). Recurring dev-testability friction.
 - **Gitea-specific quirks cost cycles:** secrets with the `GITEA_` prefix are silently dropped (→ `REGISTRY_PAT`); `actions/upload-artifact@v4` is broken on Gitea (needs the `ChristopherHX` fork); `actions/cache@v4` timed out; skipped jobs may not emit a commit-status (drove the always-running `gate` aggregate). None are documented as GitHub-incompatible up front.
 - **Scope grew mid-milestone:** the milestone planned as 7–17 but accreted 18/19/20 via `/gsd-phase`, and the ROADMAP header wasn't kept in sync — the phase-detail sections for 18–20 ended up appended after the Backlog. Keep the roadmap header + section ordering current when inserting late phases.
 ### Patterns Established
 - **Runner-probe-first** for any new self-hosted-CI capability — never author steps against an unprobed runner.
 - **Always-running `gate` aggregate** (`if: always()`, passes on success-or-skipped) is the only safe required-check surface when path-filtering jobs — never mark a path-filtered job itself required (deadlock).
 - **In-process EventEmitter over Redis** for single-process work (the outbox drain); reserve external infra for genuinely cross-process needs.
 - **Local auth is a first-class mode**, not a fallback — identity stays OIDC-`iss+sub` (never email); a local user is *linked* to an OIDC identity via an explicit claim flow (D-10/D-19).
 - **Confine dev-only affordances at build + boot:** bake `NODE_ENV=production` into the prod image and refuse-to-boot if `DEV_AUTH_BYPASS` is set — defense-in-depth beyond the runtime guard.
 ### Key Lessons
 1. Capturing deferred ideas as structured backlog entries during one milestone makes the next milestone's roadmap nearly write-itself — invest in the capture.
 2. Self-hosted GitHub-Actions-compatible runners are *not* drop-in GitHub — probe the runtime, the action ecosystem (forks), and the status/secret semantics before designing the pipeline.
 3. Dev-environment testability is a feature: if the dev user can't exercise the real flows, every feature regresses to test-only verification and the operator can't UAT — fix the dev seed/provider story early (999.19).
 ### Cost Observations
 - Model mix: not tracked
 - Sessions: not tracked
 - Notable: 14 phases shipped in ~8 days (2026-06-10 → 2026-06-18) with heavy parallelization across independent tracks (CI chain vs admin chain vs polish) once the harness landed.
 ---
 ## Cross-Milestone Trends
 ### Process Evolution
@@ -53,13 +99,17 @@ _A living document updated after each milestone. Lessons feed forward into futur
 | Milestone | Sessions | Phases | Key Change                                                                                                                    |
 | --------- | -------- | ------ | ----------------------------------------------------------------------------------------------------------------------------- |
 | v1.0      | n/a      | 6      | Established GSD plan→execute→verify→ship→complete loop; dev-auth bypass for gated infra; milestone-branch + Gitea PR shipping |
 | v1.1      | n/a      | 14     | Self-hosted Gitea CI/CD as the merge gate; per-phase branch + PR shipping; backlog→phase promotion pipeline; parallel independent tracks |
 ### Cumulative Quality
-| Milestone | Tests                                 | Coverage     | Zero-Dep Additions |
+| Milestone | Tests                                 | Coverage     | Zero-Dep Additions                          |
-| --------- | ------------------------------------- | ------------ | ------------------ |
+| --------- | ------------------------------------- | ------------ | ------------------------------------------- |
-| v1.0      | PWA 191 + API broker/events 114 green | not measured | n/a                |
+| v1.0      | PWA 191 + API broker/events 114 green | not measured | n/a                                         |
 | v1.1      | PWA ~249 + API ~347 green             | not measured | `outboxTrigger.ts` EventEmitter (CAL-15); Redis later removed entirely as unused |
 ### Top Lessons (Verified Across Milestones)
-1. (pending second milestone to cross-validate)
+1. **Capture deferred ideas as structured backlog during the milestone** — v1.1's roadmap came almost entirely from v1.0-era backlog entries.
 2. **iOS-Safari standalone / on-device push stays a human gate** across both milestones — automated harnesses (desktop + mobile-emulated) cover layout/flows, never the device-only behavior.
 3. **`setInterval` + in-process signals over external schedulers/brokers** for this single-process app — node-cron silently no-ops (v1.0), Redis went unused (v1.1).
@@ -3,7 +3,9 @@
 ## Milestones
 - ✅ **v1.0 MVP** — Phases 1–6 (shipped 2026-06-10) — see [`milestones/v1.0-ROADMAP.md`](milestones/v1.0-ROADMAP.md)
- 🚧 **v1.1 Operability & Polish** — Phases 7–17 (planning) — mobile test harness, Gitea CI (runs the harness), faster write-back, in-app admin, per-event reminders, guided setup, real lint gate, desktop e2e, doc-only CI skip + markdown lint, CI dependency audit + security checks + image hygiene, UI optimization & polish
+- ✅ **v1.1 Operability & Polish** — Phases 7–20 (shipped 2026-06-18) — see [`milestones/v1.1-ROADMAP.md`](milestones/v1.1-ROADMAP.md)
 > Next milestone not yet defined — start with `/gsd-new-milestone`.
 ## Phases
@@ -21,439 +23,52 @@ Full phase detail archived in [`milestones/v1.0-ROADMAP.md`](milestones/v1.0-ROA
 </details>
-### 🚧 v1.1 Operability & Polish (Phases 7–17)
+<details>
-
+<summary>✅ v1.1 Operability & Polish (Phases 7–20) — SHIPPED 2026-06-18</summary>
-Make FamilySync configurable, administrable, and maintainable for real multi-member use — without hand-editing env files or the database. The new critical path runs **mobile test harness → Gitea CI** (CI consumes the harness specs for UI regression), and the **admin role → reminders / setup wizard** chain (a single `/api/admin` + `/api/setup` route surface carrying the v1.1 DB migration). Faster write-back is a fully independent track.
+
-
+- [x] Phase 7: Mobile Test Harness (4/4 plans) — completed 2026-06-11
- [x] **Phase 7: Mobile Test Harness** - Mobile-emulated, authenticated PWA browser harness so the assistant (and CI) can catch mobile-only defects (completed 2026-06-11)
+- [x] Phase 8: Gitea CI (4/4 plans) — completed 2026-06-11
- [x] **Phase 8: Gitea CI** - Full regression on PR to main (lint/typecheck/unit/API-integration vs a MariaDB service container **+ the Phase 7 mobile harness as a UI-regression step against a CI-hosted dev stack**) + Docker image publish on merge (completed 2026-06-11)
+- [x] Phase 9: Faster Write-Back (2/2 plans) — completed 2026-06-12
- [x] **Phase 9: Faster Write-Back** - Event-driven outbox drain so edits land in ~1-2s instead of ~15s, preserving every outbox durability guarantee (completed 2026-06-12)
+- [x] Phase 10: Admin Role & Settings (4/4 plans) — completed 2026-06-13
- [x] **Phase 10: Admin Role & Settings** - DB foundation (is_admin / reminder_lead / app_config) + role-gated admin UI to rotate app passwords and designate the shared calendar (completed 2026-06-13)
+- [x] Phase 11: Per-Event Reminders (5/5 plans) — completed 2026-06-14
- [x] **Phase 11: Per-Event Reminders** - Reminder selector on the event form (incl. "None") serialized as VALARM, with a variable-lead scheduler that honors each event's choice (completed 2026-06-14)
+- [x] Phase 12: Initial Setup Wizard (7/7 plans) — completed 2026-06-16
- [x] **Phase 12: Initial Setup Wizard** - First-run validated bootstrap of env/VAPID/DB/OIDC + first app password, reusing the admin route surface (completed 2026-06-16)
+- [x] Phase 13: Real Lint Gate (ESLint) (3/3 plans) — completed 2026-06-12
- [x] **Phase 13: Real Lint Gate (ESLint)** - Wire ESLint flat config (typescript-eslint + React) across both apps so the Phase 8 CI lint slot actually fails on violations instead of no-op'ing (completed 2026-06-12)
+- [x] Phase 14: Desktop E2E Coverage (1/1 plans) — completed 2026-06-12
- [x] **Phase 14: Desktop E2E Coverage** - Add a Desktop Chrome Playwright profile + make the mobile-authored specs desktop-safe so the Phase 8 regression gate validates desktop, not just mobile (completed 2026-06-12)
+- [x] Phase 15: Doc-Only CI Skip + Markdown Lint (3/3 plans) — completed 2026-06-12
- [x] **Phase 15: Doc-Only CI Skip + Markdown Lint** - Aggregate-gate the slow api/harness CI jobs so doc-only PRs to main merge without running them (no branch-protection deadlock), and add markdownlint to `fast-checks` so docs get a fast format+lint gate (promoted from backlog 999.17) (completed 2026-06-12)
+- [x] Phase 16: CI Dependency Audit, Security & Image Hygiene (6/6 plans) — completed 2026-06-13
- [x] **Phase 16: CI Dependency Audit, Security Checks & Image Hygiene** - Extend Gitea CI with outdated-dependency reporting + vulnerability audit + a baseline of additional security checks, and enforce the dev/prod image boundary so no dev-bypass, secret, or family data ships in published images (absorbs backlog 999.17); independent of the admin chain (completed 2026-06-13)
+- [x] Phase 17: UI Optimization & Polish (6/6 plans) — completed 2026-06-18
- [x] **Phase 17: UI Optimization & Polish** - Phone-layout polish + branding + theme groundwork: fix the long-standing phone-layout overlap where the fixed BottomTabBar covers the New Event FAB and the colour legend (+ small-viewport sweep), finish the branding assets (real FamilySync logo into the BrandSlot seam + a complete favicon/PWA-icon set replacing the placeholder stubs), and restructure tokens.css into a themeable token layer (light-only groundwork for future dark mode). Shipped dark theme → backlog 999.20; broader styling refresh → backlog 999.21 (future milestone) (completed 2026-06-18)
+- [x] Phase 18: Auto Timezone Detection (4/4 plans) — completed 2026-06-14
-
+- [x] Phase 19: Local Auth (No-OIDC Mode) (5/5 plans) — completed 2026-06-17
-## Phase Details
+- [x] Phase 20: Admin Member Editor & Form Declutter (3/3 plans) — completed 2026-06-18
-
+
-> v1.0 phase detail (Phases 1–6) is archived in [`milestones/v1.0-ROADMAP.md`](milestones/v1.0-ROADMAP.md).
+Full phase detail archived in [`milestones/v1.1-ROADMAP.md`](milestones/v1.1-ROADMAP.md).
-
+
-### Phase 7: Mobile Test Harness
+</details>
 **Goal**: The assistant can drive the PWA in a mobile-emulated, authenticated browser context against the host-side dev stack, so mobile-only layout and flow defects can be caught automatically instead of only by the operator on real devices. This harness is also the artifact Phase 8 (CI) runs for UI regression.
 **Mode:** standard
 **Depends on**: Nothing (fully independent; goes first. One new dev dependency `@playwright/test` in `apps/pwa`; no backend changes).
 **Requirements**: TEST-01, TEST-02
 **Success Criteria** (what must be TRUE):
  1. An automated run can load the PWA in a mobile-emulated viewport (device profile + mobile UA + touch) and assert on responsive layout / tap targets.
  2. The automated run reaches the authenticated PWA via the existing `DEV_AUTH_BYPASS` on the host-side dev stack — no manual login and no Authelia/OIDC mocking.
  3. The harness runs repeatably day-over-day without re-capturing any session state (no stale storage-state failures).
  4. The harness specs are structured so they can run headlessly in CI (Phase 8) against a dev stack the runner brings up — no dependence on a developer's already-running host stack.
 **Pitfalls this phase owns** (from PITFALLS.md):
  - **No stale storage-state** (Pitfall 14): use `DEV_AUTH_BYPASS=true` for the automated harness rather than a checked-in storage-state.json with an expiring session cookie; decide the auth strategy before the first test.
  - **Service worker block** (Pitfall 15): set `serviceWorkers: 'block'` (or explicitly unregister) in the context so a previous run's SW does not intercept requests / return stale cached responses; verify no SW-sourced responses in the trace.
  - Hard constraints: targets the dev build via `DEV_AUTH_BYPASS` (DEV_AUTH_BYPASS user 1 has no CalDAV credential/calendars — verify layout/flows, not live event-create); real prod-service-worker / iOS-Safari-standalone mobile testing stays a human/device gate (out of scope).
 **Plans**: 4 plans (3 waves)Plans:
 **Wave 1**
 - [x] 07-01-PLAN.md — Harness foundation: @playwright/test + WebKit/Chromium browsers, playwright.config.ts (iPhone/WebKit + Pixel/Chromium matrix, serviceWorkers block, env baseURL, vite webServer), vitest exclude, scripts (Wave 1)
 **Wave 2** *(blocked on Wave 1 completion)*
 - [x] 07-02-PLAN.md — global-setup.ts: /health readiness poll + deterministic mysql2 reset-and-seed (calendar id 10 INSERT IGNORE guard, list + items) + e2e README/guardrails (Wave 2)
 **Wave 3** *(blocked on Wave 2 completion)*
 - [x] 07-03-PLAN.md — layout.spec.ts: tap targets >=44px, no overflow, in-viewport, accessible names (UI-SPEC Rules 1-4) + harness self-validation injected-defect proofs (Wave 3)
 - [x] 07-04-PLAN.md — calendar.spec.ts + lists.spec.ts: populated/empty/error states (Rules 4/5) + DEV_AUTH_BYPASS auth-reached + no-SW-controller precondition (Wave 3)
 **UI hint**: yes
 ### Phase 8: Gitea CI
 **Goal**: Every PR to `main` runs a full regression that gates the merge — lint, typecheck, unit, API-integration against a MariaDB service container, **and the Phase 7 mobile Playwright harness as a UI-regression step against a CI-hosted dev stack** — and a merge to `main` builds and publishes the API Docker image, all on the existing self-hosted Gitea Actions runner.
 **Mode:** standard
 **Depends on**: Phase 7 (the PR regression runs the Phase 7 mobile harness specs as its UI-regression step; without the harness there is nothing to run). No other code dependencies. Start with a runner-probe step.
 **Requirements**: CI-01, CI-02
 **Success Criteria** (what must be TRUE):
  1. Opening or updating a PR targeting `main` triggers a workflow that runs lint, typecheck (both apps), unit tests, and API integration tests against a MariaDB service container — and a failing run blocks the merge.
  2. The API integration tests connect to the service-container MariaDB (DB_HOST=127.0.0.1, service creds) and pass reliably on a cold first run, not only on re-run.
  3. The same PR workflow brings up the dev stack inside the runner — the API dev server, the PWA dev server, and the MariaDB service container, with `DEV_AUTH_BYPASS=true` — and runs the Phase 7 mobile Playwright harness specs headlessly against that authed PWA; a harness failure blocks the merge.
  4. The harness step waits for both the API and PWA dev servers to be ready (readiness probe / poll) before launching Playwright, so it does not flake on startup races.
  5. On merge to `main`, the API Docker image is built and pushed to the Gitea container registry under a sensible tag.
  6. Registry credentials never appear in plaintext in the CI logs.
 **Pitfalls this phase owns** (from PITFALLS.md):
  - **Runner-probe first** (Pitfall 12): the first workflow only probes `node --version` / `pnpm --version` / Docker access on the `self-hosted` runner before any test or build steps are designed; pin Node 22 explicitly, do not assume `actions/setup-node` works as on GitHub.
  - **MariaDB readiness wait** (Pitfall 11): add an explicit readiness loop (e.g. `healthcheck.sh --connect --innodb_initialized`, NOT `mysqladmin ping` which is removed in MariaDB 11) before any `drizzle-kit migrate` / integration test step; healthy ≠ accepting connections.
  - **Dev-stack readiness races (NEW for the harness step):** running the PWA and API dev servers *inside* CI adds startup/readiness races on top of the MariaDB-11 readiness race. The harness step must wait for **both** the API and PWA dev servers to be accepting connections (poll their URLs / health endpoints) before Playwright launches — do not race the browser against a not-yet-listening server. Run with `DEV_AUTH_BYPASS=true` so the harness reaches the authed PWA exactly as in Phase 7.
  - **--password-stdin** (Pitfall 13): `docker login` via `--password-stdin` with the token piped from a registered Gitea secret (PAT with `write:package`); never `-p $TOKEN` on the command line.
  - Hard constraints: API integration tests need a real MariaDB and live in `apps/api/tests/` (never `src/`); cache the pnpm store; Drizzle generate+migrate to set up the CI DB schema; the harness step reuses the Phase 7 specs unchanged (CI owns only the stack bring-up + readiness wait, not the spec content).
 **Plans**: 4 plans (4 waves)Plans:
 **Wave 1**
 - [x] 08-01-PLAN.md — Runner probe + operator runner/PAT registration (W0; answers the Docker-vs-host fork)
 **Wave 2** *(blocked on Wave 1 completion)*
 - [x] 08-02-PLAN.md — ci.yml: fast-checks (lint/typecheck/PWA unit) + API job (MariaDB service + migrate + DB-backed tests)
 **Wave 3** *(blocked on Wave 2 completion)*
 - [x] 08-03-PLAN.md — ci.yml: harness job (dev-stack bring-up + readiness waits + Phase 7 Playwright specs, both profiles)
 **Wave 4** *(blocked on Wave 3 completion)*
 - [x] 08-04-PLAN.md — ci.yml: publish job (build production image + push :latest + :v1.1-<sha> via --password-stdin)
 **UI hint**: yes
 ### Phase 9: Faster Write-Back
 **Goal**: A created, edited, or deleted event reaches Fastmail within ~1-2 seconds (event-driven outbox drain) instead of waiting up to ~15s for the next interval tick — with every existing durability guarantee intact.
 **Mode:** standard
 **Depends on**: Nothing (fully independent; the only new artifact is a zero-dependency in-process EventEmitter, `lib/outboxTrigger.ts`). Can run in parallel with any other v1.1 track.
 **Requirements**: CAL-15
 **Success Criteria** (what must be TRUE):
  1. After creating/editing/deleting an event, the change lands in Fastmail in ~1-2s in the common case (drain is signalled on enqueue, not waited-for on the interval) — observable as the change appearing in the Fastmail native app well before the old ~15s window.
  2. The route handler still returns an optimistic 202 immediately and never makes a CalDAV call inline — the event-driven signal is fire-and-forget.
  3. Edit-as-move still writes the new event before deleting the old one (create-before-delete ordering preserved); no event is ever lost when a move drains under rapid enqueues.
  4. No duplicate CalDAV PUTs occur for the same outbox row when the signal and the 15s fallback interval overlap (exactly-once per uid preserved).
  5. The 15s `setInterval` fallback still runs and recovers any rows missed by the signal path (startup catch-up, transient errors).
 **Pitfalls this phase owns** (from PITFALLS.md):
  - **No double-drain** (Pitfall 5): the trigger must set a `drainRequested` flag funnelled through the single setInterval-controlled path / the existing `isDraining` guard — never call `runOutboxDrain()` directly from the signal in a way that bypasses the guard or escapes the error-caught wrapper.
  - **Create-before-delete under concurrent enqueues** (Pitfall 6): enqueue CREATE before DELETE; do not fire the signal between the two inserts of a move (publish after both inserts / after the transaction commits).
  - Hard constraints: `setInterval` only (no node-cron); single-process by design — **no Redis** for the drain (Redis stays for list SSE); all outbox guarantees (fresh-etag-before-PUT, 412 conflict flow, per-uid exactly-once) unchanged.
 **Plans**: 2 plans (2 waves)
 Plans:
 **Wave 1**
 - [x] 09-01-PLAN.md — TDD: outboxTrigger.ts (zero-dep EventEmitter signal) + scheduleOutboxDrain wrapper / drainRequested trailing-re-drain loop + initOutboxTrigger in outboxWorker.ts; trigger-wiring tests (SC-1, SC-4/D-07, D-05) (Wave 1)
 **Wave 2** *(blocked on Wave 1 completion)*
 - [x] 09-02-PLAN.md — Four post-commit signalOutboxDrain() publish sites in events.ts (create / edit-as-move-after-transaction / same-cal update / delete) + initOutboxTrigger() startup wiring under isMainModule() in index.ts (Wave 2)
 ### Phase 10: Admin Role & Settings
 **Goal**: An admin can manage household configuration that previously required manual DB writes — rotating a member's Fastmail app password and designating the shared family calendar — from a role-gated in-app Settings section, on top of the v1.1 DB foundation this phase introduces.
 **Mode:** standard
 **Depends on**: Nothing required upstream; this phase **carries the v1.1 DB migration** (users.is_admin, calendar_events.reminder_lead_minutes, app_config table) that Phases 11 and 12 build on. It is the head of the admin chain (10 → 11, 10 → 12).
 **Requirements**: ADMIN-01, ADMIN-02, ADMIN-03
 **Success Criteria** (what must be TRUE):
  1. An admin sees an Admin section in Settings and can list household members with their credential status; a non-admin member never sees it and cannot invoke any `/api/admin/*` route (gets 403).
  2. An admin can enter or rotate a member's Fastmail app password; it is validated against CalDAV (PROPFIND) before saving and stored encrypted — and the password is never displayed, echoed in a response, or logged.
  3. An admin can pick which synced calendar is the shared family calendar from a list, and the `calendars.is_shared` flag updates accordingly (replacing the manual `UPDATE calendars SET is_shared=1` step).
  4. The role check is role-agnostic and member-count-agnostic: it gates on `users.is_admin`, so more admins can be added later without reworking the guard.
  5. The DB migration (is_admin, reminder_lead_minutes, app_config) is applied via generate+migrate and is in place for downstream phases (reminder_lead_minutes for Phase 11, app_config.setup_complete for Phase 12).
 **Pitfalls this phase owns** (from PITFALLS.md):
  - **Admin role check inside the sub-router** (Pitfall 9): apply `requireAdmin` with `.use('*', ...)` inside `adminRouter`, not only at the parent mount; integration test must assert 403 for a non-admin authenticated user.
  - **App password never logged/echoed** (Pitfall 7): custom zod-validator `hook` returns a generic 400 (no Zod `received`/`value` field); no `console.log` of request bodies in `routes/admin*`.
  - Hard constraints: Drizzle **generate+migrate, never push** (false destructive diff on populated MariaDB); reuse `broker/crypto.ts` `encryptPassword` (no changes to crypto); `/api/admin/credentials` and `/api/admin/calendars/:id/shared` are the single shared surface — do NOT duplicate them into `/api/setup/*` in Phase 12.
 **Folded-in scope** (from backlog 999.5, self-service member onboarding): the credential surface this phase builds is the same one a member needs on first login. Expose a `needsProviderSetup` signal (member has no `member_credentials` row) and let a member enter/validate (CalDAV PROPFIND) + encrypt their **own** Fastmail app password — the self-service counterpart of the admin-managed flow, sharing the validation/encryption/initial-sync path. Non-technical-friendly instructions (link to Fastmail's app-password page, required Calendars/CalDAV scope) are the hard UX constraint. Member-scoped: a member can only set their own credential; never log/echo the password.
 **Plans**: 4 plans (4 waves)Plans:
 **Wave 1**
 - [x] 10-01-PLAN.md — v1.1 DB foundation migration (is_admin, provider_type+unique, reminder_lead_minutes, app_config) + dev-bypass admin seed
 **Wave 2** *(blocked on Wave 1 completion)*
 - [x] 10-02-PLAN.md — requireAdmin guard + first-login-wins bootstrap + /api/me isAdmin/needsProviderSetup (TDD)
 **Wave 3** *(blocked on Wave 2 completion)*
 - [x] 10-03-PLAN.md — adminRouter (members/credentials/calendars/shared) + member self-service credential, validate→encrypt→sync (TDD)
 **Wave 4** *(blocked on Wave 3 completion)*
 - [x] 10-04-PLAN.md — PWA /admin route + nav gating + CredentialSheet + SetupBanner (playwright-cli verified)
 **UI hint**: yes
 ### Phase 11: Per-Event Reminders
 **Goal**: A user can choose a reminder lead time per event (None / 5m / 10m / 15m / 30m / 1h / 2h / 1d / 2d, default None), serialized as a VALARM on the event, and the push scheduler fires at that exact lead — firing nothing when there is no alarm and never stripping reminders set in other clients.
 **Mode:** standard
 **Depends on**: Phase 10 (the `calendar_events.reminder_lead_minutes` column from the v1.1 migration is the scheduler's ground truth). Independent of Phases 7/8/9/12.
 **Requirements**: CAL-13, CAL-14, NOTIF-04, NOTIF-05, NOTIF-06
 **Success Criteria** (what must be TRUE):
  1. When creating or editing a timed event, the user can pick a reminder lead from the preset list (None default); the choice round-trips to Fastmail as a VALARM and is visible/honored on re-open.
  2. Editing an event that already has a reminder set in another client (Fastmail / Apple Calendar) preserves that VALARM — it is never silently dropped on round-trip.
  3. A reminder push fires at the event's chosen lead time (e.g. T-30 for a 30-minute lead), not a hardcoded 15-minute lead.
  4. An event with no reminder set produces no reminder push (no default 15-minute fire).
  5. An all-day event's reminder fires at a sensible local time (9 AM on the alert day), not midnight; the reminder selector is disabled/hidden for all-day events in the UI; and reminder delivery stays exactly-once across catch-up scans and rescheduled events.
 **Pitfalls this phase owns** (from PITFALLS.md):
  - **Preserve-on-edit** (Pitfall 1): the update path extracts and preserves existing VALARM sub-components from `rawVevent` (mirroring the WR-01 RRULE-preserve pattern) — never rebuild-from-scratch and silently strip; `outboxPayloadSchema` distinguishes "no change" from explicit "no reminder".
  - **No TRIGGER VALUE=TEXT** (Pitfall 2): build the trigger with `ICAL.Duration.fromSeconds(-n*60)`, not a bare string; unit-test that the ICS emits a DURATION trigger with no `VALUE=TEXT`.
  - **All-day 9AM semantics** (Pitfall 3): guard `buildVeventString` (`if (!allDay && reminderMinutes > 0)`), disable the selector when allDay, keep the scheduler's all-day handling at 9 AM local.
  - **uid:dtstartMs dedup** (Pitfall 4): change the scheduler dedup key from bare `uid` to compound `uid:dtstartMs` and widen the scan to a variable per-event window so long leads fire and rescheduled events re-fire; keep `eventFieldsSchema` and `outboxPayloadSchema` in sync (IN-03).
  - Hard constraints: `setInterval` only; scheduler reads `reminder_lead_minutes` from the DB (ground truth), not the outbox payload; drop the `isShared`-only reminder restriction (a user who set an alarm wants it regardless of calendar).
 **Plans**: 4 plans (3 waves)
 Plans:
 **Wave 1**
 - [x] 11-01-PLAN.md — VALARM builders + classifier + extractor + computeAlertInstantUtc (vevent.ts, TDD)
 **Wave 2** *(blocked on Wave 1 completion)*
 - [x] 11-02-PLAN.md — Variable-lead scheduler: uid:dtstartMs dedup, drop isShared, all-day 9 AM, humanized body (TDD)
 - [x] 11-03-PLAN.md — Backend plumbing: schema field, outbox preserve-on-edit, sync upsert, occurrence surfacing
 **Wave 3** *(blocked on Wave 2 completion)*
 - [x] 11-04-PLAN.md — EventForm reminder picker (allDay swap, edit pre-population) + client types + Playwright smoke
 **UI hint**: yes
 ### Phase 12: Initial Setup Wizard
 **Goal**: On first run (no admin/credentials configured), the operator is guided through a validated, step-by-step wizard to bootstrap the app — env presence, generated secrets to copy, DB/OIDC/VAPID/app-password validation — instead of hand-editing `.env` / `docker-compose.yml`; once complete, the setup endpoints lock.
 **Mode:** standard
 **Depends on**: Phase 10 (reuses the admin role + `/api/admin/credentials` and `/api/admin/calendars/:id/shared` routes; the wizard is the second frontend consumer of that surface, and `app_config` from the Phase 10 migration holds `setup_complete`). Goes last. Independent of Phases 7/8/9/11.
 **Requirements**: SETUP-01, SETUP-02, SETUP-03, SETUP-04
 **Success Criteria** (what must be TRUE):
  1. On a fresh install with nothing configured, the operator reaches a setup wizard (via `GET /api/setup/status` mounted before the OIDC guard) and walks through bootstrap steps instead of editing files by hand.
  2. Each input is validated before the step can complete: DB connects, VAPID private key decodes to exactly 32 bytes and pairs with the public key, OIDC discovery resolves, and the Fastmail app password reaches CalDAV (PROPFIND).
  3. Generated secrets (session secret, encryption key, VAPID keypair) are displayed for the operator to copy into env; they are never written to the DB or returned in a way that persists, and `APP_PASSWORD_ENCRYPTION_KEY`/`VAPID_PRIVATE_KEY` never enter the DB at all.
  4. After completion, the wizard-completing user is promoted to admin (`is_admin`), `app_config.setup_complete` is set, and any further call to a setup endpoint returns 423 Locked.
  5. The 423 guard is enforced on every invocation (checked against member-credentials + VAPID env present), not only at startup.
 **Pitfalls this phase owns** (from PITFALLS.md):
  - **Guard on every invocation** (Pitfall 8): the "already set up" guard returns 423 from all setup routes once configured — implement and test the guard before the happy path; a second POST after completion must return 423, not 200.
  - **Secrets stay in env, never in DB** (Pitfalls 8 & 10): the wizard validates secrets by performing a test operation (test encrypt/decrypt, structural VAPID check), never by accepting/storing the key value; no DB column for `vapid_private_key` or `app_password_encryption_key`; never log/echo the app password.
  - Hard constraints: `GET /api/setup/status` mounts **before** the OIDC guard (like `/health`); do NOT create `/api/setup/credentials` — reuse the Phase 10 admin routes; Drizzle generate+migrate (any `app_config` seeding via migration).
 **Plans**: 7 plans in 4 waves (4 original + 3 gap-closure for 12-UAT.md gaps 1-6)
 Plans:
 **Wave 1**
 - [x] 12-01-PLAN.md — Schema migration (nullable OIDC + claimed) + generate-secrets helper (SETUP-03) + Wave-0 scaffolds
 **Wave 2** *(blocked on Wave 1 completion)*
 - [x] 12-02-PLAN.md — Pre-auth /api/setup/* router + isSetupLocked 423 guard + index mount + OIDC boot fallback (SETUP-01/02/04)
 - [x] 12-03-PLAN.md — First-login-claims rework in upsertUser (D-08, SETUP-01)
 **Wave 3** *(blocked on Wave 2 completion)*
 - [x] 12-04-PLAN.md — PWA SetupPage wizard + App.tsx gate + UI-SPEC revision (SETUP-01/02)
 **Wave 4 — Gap closure** *(UAT 12-UAT.md gaps 1-6; 06+07 parallel, 05 blocked on 06)*
 - [x] 12-06-PLAN.md — Backend: validate/vapid asserts wizard key == env VAPID_PUBLIC_KEY (gap 2) + status exposes non-secret DB name (gap 3) (SETUP-02)
 - [x] 12-07-PLAN.md — App.tsx: reverse-gate /setup post-completion (gap 5) + reconcile ['me'] so calendar banner clears after wizard (gap 6) (SETUP-01/04)
 - [x] 12-05-PLAN.md — SetupPage: drop DB-vs-env aside (gap 1) + read-only DB-name field (gap 3) + persist fields across Back (gap 4) (SETUP-01) — depends on 12-06
 **UI hint**: yes
 ### Phase 13: Real Lint Gate (ESLint)
 **Goal**: The CI lint gate actually fails on lint violations. A real ESLint flat config (`eslint.config.js`, `typescript-eslint`; React + react-hooks plugins for `apps/pwa`) plus a package-level `lint` script in `apps/api` and `apps/pwa` makes the existing root `pnpm -r --if-present lint` run a real linter, replacing the hollow no-op gate that exits 0 because no linter exists.
 **Mode:** standard
 **Depends on**: Phase 8 (the CI `fast-checks` job already runs `pnpm lint`; this fills the slot Phase 8 shipped wired to auto-activate once a package `lint` script lands). Independent of all other phases.
 **Requirements**: TBD (promoted from backlog 999.16)
 **Success Criteria** (what must be TRUE):
  1. `pnpm lint` runs ESLint across both `apps/api` and `apps/pwa` and exits non-zero on an introduced violation (verified by a deliberate test violation), where today it exits 0 with no linter present.
  2. The CI `fast-checks` lint step blocks a PR to main on lint violations — the gate can now fail.
  3. The first real run's existing violations are resolved (fix / warn / disable decided per rule) so the baseline gate ends green.
 **Pitfalls this phase owns**:
  - Pick a baseline ruleset (recommended vs strict-type-checked) deliberately — strict surfaces a large upfront cleanup; decide blocking vs advisory before flipping the gate to blocking.
  - `typecheck`/tsc already gates type errors; ESLint should not duplicate type-checking rules unnecessarily.
 **Plans**: 3 plans — all complete (scope expanded during planning to add a Prettier `format:check` gate)
 - [x] 13-01-PLAN.md — Install ESLint/Prettier deps + flat config + package scripts + prove the gate fails (SC-1)
 - [x] 13-02-PLAN.md — Fix all first-run lint violations across both apps, green `pnpm lint` (D-13-05/06)
 - [x] 13-03-PLAN.md — Prettier reformat (isolated commit) + CI format:check step + green baseline (SC-2/SC-3)
 **UI hint**: no
 ### Phase 14: Desktop E2E Coverage
 **Goal**: The Phase 8 regression gate exercises the desktop layout and flows, not just mobile. A `desktop` Playwright project (`devices['Desktop Chrome']`, no touch, wide viewport) is added to `apps/pwa/playwright.config.ts`, and the existing mobile-authored specs are reviewed/adjusted (or appropriately skipped) so `pnpm test:e2e` passes on a no-touch desktop viewport as well as the `iphone`/`pixel` profiles.
 **Mode:** standard
 **Depends on**: Phase 7 (the harness it extends) and Phase 8 (CI runs `pnpm test:e2e` and picks up the new project automatically — no CI plumbing change needed beyond any desktop-profile runtime/wait). Independent of Phases 9–13.
 **Requirements**: TBD (promoted from backlog 999.15)
 **Success Criteria** (what must be TRUE):
  1. A `desktop` project exists in `playwright.config.ts` (Desktop Chrome, wide viewport, no `hasTouch`).
  2. The existing e2e specs pass (or are explicitly, justifiably skipped) on the desktop profile — touch-gesture / mobile-drawer / mobile-only-layout assumptions are handled.
  3. `pnpm test:e2e` in CI runs and gates on both mobile and desktop profiles (blocking-vs-advisory for desktop decided when planned).
 **Pitfalls this phase owns**:
  - The real work is the spec-compat pass, not CI plumbing — Phase 8 reused the Phase 7 harness unchanged, so the config addition is small but specs authored for touch/mobile need per-spec review.
  - Desktop WebKit is optional — the Apple member is already covered on mobile Safari via `iphone`; Desktop Chrome is likely sufficient for a shared/wall browser.
 **Plans**: 1 plan
 Plans:
 - [x] 14-01-PLAN.md — Add the `desktop` Playwright project, desktop-skip the two mobile-only layout assertions (+ D-04 parity), update spec/README docs, and prove `pnpm test:e2e` is green on iphone + pixel + desktop with a blocking CI gate.
 **UI hint**: no
 ### Phase 15: Doc-Only CI Skip
 **Goal**: Doc-only PRs to `main` merge without running the slow `harness` (Playwright e2e + dev-stack bring-up, ~5 min) and `api` (MariaDB integration) jobs, while `fast-checks` (Prettier `format:check` + markdown linting) still runs — and branch protection never deadlocks on a required check that never reports. Docs get a *fast but real* gate: format + lint, none of the slow code jobs.
 **Mode:** standard
 **Depends on**: Phase 8 (the `.gitea/workflows/ci.yml` it modifies) and Phase 13 (the `fast-checks` job + `format:check` step this extends). Independent of Phases 9–12.
 **Requirements**: TBD (promoted from backlog 999.17)
 **Success Criteria** (what must be TRUE):
  1. A doc-only PR to `main` (only `docs/` or `*.md` changed) skips the `api` and `harness` jobs but still runs `fast-checks`.
  2. A PR touching code runs `fast-checks`, `api`, and `harness` as today; a failure in any blocks the merge.
  3. Branch protection requires `CI / fast-checks` + an always-running `CI / gate` aggregate (passes when each heavy job is `success` OR `skipped`) — the direct `api`/`harness` requirements are dropped so a skipped heavy job never deadlocks the merge.
  4. `fast-checks` runs a markdown linter (markdownlint-cli2) over `**/*.md`; an introduced markdown-lint violation fails the gate, and the existing markdown baseline passes (violations fixed or rules configured) so the gate starts green.
 **Pitfalls this phase owns**:
  - **Required-check deadlock** — never path-filter a required context directly; a required job that never reports blocks the PR forever. The always-running `gate` job (`if: always()`, passes on `success`/`skipped`) is the only safe gating surface.
  - **Gitea skipped-status quirk** — Gitea may not emit a commit-status for a `skipped` job; rely on the always-running `gate`, not on marking `api`/`harness` skipped-but-required.
  - **Prettier vs markdownlint overlap** — Prettier already owns markdown *formatting*; scope markdownlint to *content* rules (heading increments, no broken/duplicate link refs, list/code-fence conventions) and disable its purely-stylistic rules that fight Prettier (e.g. line-length, list-indent), so the two don't conflict on the same `.md`.
  - **`.planning/*` is push-direct, never linted** — planning bookkeeping bypasses CI via the Unprotected file pattern, so markdownlint never sees it; scope the lint glob to `docs/` + repo-root/app `*.md` and exclude `.planning/**` (and any generated markdown) to avoid a baseline cleanup of churny bookkeeping files.
 **Plans**: 3 plans (3 waves)
 Plans:
 **Wave 1**
 - [x] 15-01-PLAN.md — markdownlint-cli2 + `.markdownlint-cli2.jsonc` + `md:lint` script + fast-checks step + fix 13 baseline violations (SC-4)
 **Wave 2** *(blocked on Wave 1 completion)*
 - [x] 15-02-PLAN.md — ci.yml: `changes` (dorny/paths-filter@v4) + conditional api/harness + always-running `gate` aggregate (SC-1/SC-2, SC-3 YAML)
 **Wave 3** *(blocked on Wave 2 completion)*
 - [x] 15-03-PLAN.md — operator branch-protection checkpoint (require `CI / fast-checks` + `CI / gate`, drop api/harness) + publish.yml comment update (SC-3)
 **UI hint**: no
 ### Phase 16: CI Dependency Audit, Security Checks & Image Hygiene
 **Goal**: The CI pipeline surfaces outdated and vulnerable dependencies, runs a baseline of additional security checks, and enforces a clean dev↔prod boundary in the images it publishes — so the two-person household app doesn't silently rot on stale/CVE-bearing packages, and no dev-only affordance, secret, or family-specific data ever ships in a production image. Extends the existing Gitea CI (Phase 8) workflow with dependency/security/image-hygiene gates rather than standing up a separate pipeline. **Absorbs backlog 999.17 (dev/prod image boundary).**
 **Mode:** standard
 **Depends on**: Phase 8 (Gitea CI — adds steps to the existing workflow + publish job; no admin-chain dependency). Independent of Phases 10–12.
 **Requirements**: SEC-01 (secret scanning), SEC-02 (static security lint), DEP-01 (vuln audit gate), DEP-02 (outdated advisory), IMG-01 (NODE_ENV+boot-guard), IMG-02 (.dockerignore), IMG-03 (publish image-hygiene assertions), CI-03 (security job + gate wiring)
 **Candidate scope (to be sharpened in `/gsd-discuss-phase 16`):**
  - **Outdated dependencies:** a CI step that reports dependencies behind their latest (e.g. `pnpm outdated -r`), surfaced on the PR. Decide gating vs advisory, and how to handle the pinned-version table in CLAUDE.md (the stack pins exact versions — "outdated" must not fight intentional pins).
  - **Vulnerability audit:** `pnpm audit` (or equivalent) against the lockfile, failing on a chosen severity threshold (e.g. high/critical). Decide the threshold and an allowlist/waiver mechanism for unfixable transitive advisories.
  - **Additional security checks (user is open to these — pick a sensible baseline, avoid over-build):** candidates — secret scanning on the diff (gitleaks/trufflehog), a CodeQL/`eslint-plugin-security` static pass, dependency-review on PRs, Dockerfile/image scan (e.g. trivy) of the published image.
  - **Dev/prod boundary definition & enforcement (from 999.17):** the `DEV_AUTH_BYPASS` concept (and any dev-only affordance) must be provably confined to local dev — never to production, never baked into published images. Today the guard is runtime-only (`NODE_ENV !== 'production' && DEV_AUTH_BYPASS === 'true'` in `apps/api/src/auth/devBypass.ts`); add (a) explicit documentation of what "dev image" vs "shipped image" means, and (b) build-time / boot-time enforcement (a `production` image refuses to boot — or the build aborts — if dev-bypass is enabled) as defense-in-depth.
  - **No data/secrets in published images (from 999.17):** audit the Dockerfile(s) + the Phase 8 publish job (`publish.yml`) to confirm `.env`, dev seed SQL, local DB dumps, encryption keys, OIDC secrets, the `DEV_USER` seed, and family-specific fixtures are `.dockerignore`d and never `COPY`'d. Add a CI assertion that fails the publish if a dev-bypass code path is active, a forbidden env/secret is present, or personal/seed data is staged into the image context. The dev-stack seed path (`DEV_USER` id 1 + sample calendar/list data) must be unreachable from the production image/compose.
  - **Noise control:** these gates are notorious for flaky/advisory-churn failures; decide blocking-on-merge vs warn-only per check, and where results surface (PR annotation vs job log), mirroring Phase 15's gate-aggregation approach.
 **Boundary:** Extends the existing Gitea CI workflow + publish job; does not remove dev-bypass (still needed for local verification and the Phase 7/8 harness) and does not add a new external service or a runtime dependency to the app. Automated dependency *upgrades* (e.g. Renovate/Dependabot bots) are a separate concern — decide in discuss whether they're in scope or deferred.
 **Plans**: 6 plans in 2 waves
 Plans:
 **Wave 1**
 - [x] 16-01-PLAN.md — Image-hygiene runtime: bake NODE_ENV=production + boot-time refuse-to-boot guard (IMG-01)
 - [x] 16-02-PLAN.md — pnpm audit gate + waiver allowlist + advisory-only tiered outdated report (DEP-01, DEP-02)
 - [x] 16-03-PLAN.md — Fold eslint-plugin-security into the lint gate as blocking errors + triage (SEC-02)
 - [x] 16-04-PLAN.md — gitleaks config + full-history baseline + .dockerignore (SEC-01, IMG-02)
 **Wave 2** *(blocked on Wave 1 completion)*
 - [x] 16-05-PLAN.md — Add the security job to ci.yml (gitleaks always; audit/outdated code-gated) + gate wiring (CI-03)
 - [x] 16-06-PLAN.md — publish.yml static image-hygiene assertion + boot-smoke before push (IMG-03)
 **UI hint**: no
 ### Phase 17: UI Optimization & Polish
 **Goal**: A visual-identity & polish pass for the PWA spanning three workstreams: **(A) phone-layout polish** so the phone (≤767px) layout has no fixed-chrome overlap and small-viewport spacing reads cleanly — starting with the long-standing BottomTabBar overlap that hides the New Event FAB and the colour legend, plus a small-viewport sweep; **(B) branding assets** — generate a real FamilySync logo into the existing `BrandSlot` seam (`apps/pwa/src/components/BrandSlot.tsx`) and a complete favicon/PWA-icon set replacing the placeholder stubs in `apps/pwa/public/`; **(C) theme-token groundwork** — restructure `apps/pwa/src/styles/tokens.css` into a themeable semantic-token layer (swappable by `data-theme`/`prefers-color-scheme`), light staying the only shipped theme, so a future dark theme is cheap.
 **Mode:** standard
 **Depends on**: Nothing structural (CSS/layout + assets only). Best sequenced after Phase 10 merges (the BottomTabBar gained an Admin tab and the new SetupBanner adds top pressure on phone), but otherwise independent of the admin chain.
 **Requirements**: No REQ-IDs — decisions D-01…D-10 (17-CONTEXT.md) stand in. Coverage: D-01/D-02 (A, phone overlap+guard) → 17-03; D-03/D-04 (B, assets) → 17-02; D-04/D-05 (B, wiring) → 17-04; D-06 (C, token groundwork) → 17-01; D-07/D-09 (D, logout+sheet centering) → 17-05; D-08/D-09/D-10 (D, admin toasts+reset-sheet+two-tab nav) → 17-06.
 **Scope boundary (set in `/gsd-discuss-phase 17`, 2026-06-17):** Workstream C ships token groundwork **only** — no dark palette, no theme toggle (→ backlog **999.20**). A broader "modern styling" visual refresh is **out of scope** and routed to backlog **999.21** (future milestone). Keep Phase 17 a focused polish + branding + groundwork pass, not a redesign.
 **Seed defect — phone-layout bottom-bar overlap (documented 2026-06-13; long-standing, NOT introduced by Phase 10 — the BottomTabBar dates to Phase 04):**
 At ≤767px (`window.matchMedia('(max-width: 767px)')` in `apps/pwa/src/App.tsx`) the layout switches to a 48px top AppNav + a `position: fixed` BottomTabBar (`height: calc(56px + env(safe-area-inset-bottom))`, z-index 200; `apps/pwa/src/components/BottomTabBar.tsx`) + a floating "New Event" FAB (`position: fixed; bottom: var(--space-6); right: var(--space-6)`; `apps/pwa/src/components/CalendarShell.tsx`). Two problems:
  1. **FAB sits inside the bar** — the FAB's `bottom` offset (~`--space-6`, ≈24px) is smaller than the bar's 56px height, so the round New Event button overlaps the bottom tab bar (lands on the Admin tab).
  2. **Content occluded** — the content area (`contentStyle` in `App.tsx`) reserves no `padding-bottom` for the fixed bar, so the bottom of the calendar and the colour-legend chips (e.g. the "Dev User" / member legend) slide under the bar and are partially hidden.
 **Fix sketch (CSS-only, no behaviour change):** on phone, lift the FAB to `bottom: calc(56px + env(safe-area-inset-bottom, 0px) + var(--space-6))` and add a matching `padding-bottom: calc(56px + env(safe-area-inset-bottom, 0px))` to the phone content/scroll area (or reduce the `100dvh` column by the bar height). Verify across the `iphone`/`pixel`/`desktop` Playwright profiles and a real narrow Chromium via playwright-cli.
 **Evidence:** reproduced 2026-06-13 with playwright-cli at 390×844 (FAB over the Admin tab; "Dev User" legend clipped) vs 1280×800 (desktop sidebar, no overlap). Full detail in todo `2026-06-13-pwa-phone-bottombar-overlap.md`.
 **Candidate scope (to sharpen in `/gsd-discuss-phase 17`):** the seed defect above, plus a sweep for other small-viewport spacing / tap-target / overlap issues (the Phase 7 `layout.spec.ts` tap-target/overflow assertions are a ready checklist) and any phone/desktop visual inconsistencies noticed in use. Keep it a focused polish pass, not a redesign.
 **Plans**: 6/6 plans complete
 Plans:
 **Wave 1**
 - [x] 17-01-PLAN.md — C: tokens.css themeable-layer groundwork + --bottom-chrome-h token (D-06) [Wave 1]
 - [x] 17-02-PLAN.md — B: generate logo + full icon set, operator approval checkpoint (D-03, D-04) [Wave 1]
 **Wave 2** *(blocked on Wave 1 completion)*
 - [x] 17-03-PLAN.md — A: phone FAB/BottomTabBar overlap fix + sweep + CI overlap assertion (D-01, D-02) [Wave 2, dep 01]
 - [x] 17-04-PLAN.md — B: wire logo into BrandSlot + index.html favicons + manifest maskable + accent (D-04, D-05) [Wave 2, dep 01,02]
 - [x] 17-05-PLAN.md — D: logout control + sheet desktop-centering (SettingsSheet/CredentialSheet) (D-07, D-09) [Wave 2, dep 01]
 - [x] 17-06-PLAN.md — D: admin success toasts + two-tab ARIA nav + reset-sheet centering + admin.spec.ts (D-08, D-09, D-10) [Wave 2, dep 01]
 **UI hint**: yes
 ## Progress
 | Phase | Milestone | Plans Complete | Status   | Completed  |
 | ----- | --------- | -------------- | -------- | ---------- |
-| 1. Foundation + Broker Spike     | v1.0 | 4/4   | Complete | 2026-06-04 |
+| 1. Foundation + Broker Spike      | v1.0 | 4/4   | Complete | 2026-06-04 |
-| 2. Calendar Display              | v1.0 | 5/5   | Complete | 2026-06-05 |
+| 2. Calendar Display               | v1.0 | 5/5   | Complete | 2026-06-05 |
-| 3. Event Write-Back + PWA Install| v1.0 | 12/12 | Complete | 2026-06-07 |
+| 3. Event Write-Back + PWA Install | v1.0 | 12/12 | Complete | 2026-06-07 |
-| 4. Shared Lists + Live Sync      | v1.0 | 7/7   | Complete | 2026-06-09 |
+| 4. Shared Lists + Live Sync       | v1.0 | 7/7   | Complete | 2026-06-09 |
-| 5. Web Push Notifications        | v1.0 | 8/8   | Complete | 2026-06-10 |
+| 5. Web Push Notifications         | v1.0 | 8/8   | Complete | 2026-06-10 |
-| 6. UX Polish                     | v1.0 | 6/6   | Complete | 2026-06-10 |
+| 6. UX Polish                      | v1.0 | 6/6   | Complete | 2026-06-10 |
-| 7. Mobile Test Harness           | v1.1 | 4/4 | Complete    | 2026-06-11 |
+| 7. Mobile Test Harness            | v1.1 | 4/4   | Complete | 2026-06-11 |
-| 8. Gitea CI                      | v1.1 | 4/4 | Complete   | 2026-06-11 |
+| 8. Gitea CI                       | v1.1 | 4/4   | Complete | 2026-06-11 |
-| 9. Faster Write-Back             | v1.1 | 2/2 | Complete    | 2026-06-12 |
+| 9. Faster Write-Back              | v1.1 | 2/2   | Complete | 2026-06-12 |
-| 10. Admin Role & Settings        | v1.1 | 4/4 | Complete    | 2026-06-13 |
+| 10. Admin Role & Settings         | v1.1 | 4/4   | Complete | 2026-06-13 |
-| 11. Per-Event Reminders          | v1.1 | 5/5 | Complete    | 2026-06-14 |
+| 11. Per-Event Reminders           | v1.1 | 5/5   | Complete | 2026-06-14 |
-| 12. Initial Setup Wizard         | v1.1 | 7/7 | Complete    | 2026-06-16 |
+| 12. Initial Setup Wizard          | v1.1 | 7/7   | Complete | 2026-06-16 |
-| 13. Real Lint Gate (ESLint)      | v1.1 | 3/3   | Complete    | 2026-06-12 |
+| 13. Real Lint Gate (ESLint)       | v1.1 | 3/3   | Complete | 2026-06-12 |
-| 14. Desktop E2E Coverage         | v1.1 | 1/1 | Complete    | 2026-06-12 |
+| 14. Desktop E2E Coverage          | v1.1 | 1/1   | Complete | 2026-06-12 |
-| 15. Doc-Only CI Skip + MD Lint   | v1.1 | 3/3 | Complete    | 2026-06-12 |
+| 15. Doc-Only CI Skip + MD Lint    | v1.1 | 3/3   | Complete | 2026-06-12 |
-| 16. CI Dep Audit, Sec & Img Hyg  | v1.1 | 6/6 | Complete    | 2026-06-13 |
+| 16. CI Dep Audit, Sec & Img Hyg   | v1.1 | 6/6   | Complete | 2026-06-13 |
-| 17. UI Optimization & Polish     | v1.1 | 6/6 | Complete    | 2026-06-18 |
+| 17. UI Optimization & Polish      | v1.1 | 6/6   | Complete | 2026-06-18 |
-| 18. Auto Timezone Detection      | v1.1 | 4/4 | Complete    | 2026-06-14 |
+| 18. Auto Timezone Detection       | v1.1 | 4/4   | Complete | 2026-06-14 |
-| 19. Local Auth (No-OIDC Mode)    | v1.1 | 5/5 | Complete    | 2026-06-17 |
+| 19. Local Auth (No-OIDC Mode)     | v1.1 | 5/5   | Complete | 2026-06-17 |
-| 20. Admin Member Editor & Declutter | v1.1 | 3/3 | Complete    | 2026-06-18 |
+| 20. Admin Member Editor & Declutter | v1.1 | 3/3 | Complete | 2026-06-18 |
 ## Backlog
@@ -696,77 +311,3 @@ Plans:
 - [ ] TBD (promote with /gsd-review-backlog when ready)
 ### Phase 18: Auto timezone detection and ability to change timezone
 **Goal:** Make the household timezone an explicit, stored, user-changeable setting — auto-detected from the browser at first run, changeable from the role-gated /admin Settings — and route the server-side all-day "9 AM local" reminder computation through it (replacing the implicit `process.env.TZ` fallback), without touching the already-correct browser-local display/timed-write path.
 **Requirements**: TBD (decision contract D-01..D-07 from 18-CONTEXT.md)
 **Depends on:** Phase 10 (admin role + `/admin` Settings + `app_config`); Phase 11 (all-day reminder computation this rewires). Independent of Phase 17. Phase 12 (setup wizard) not required — seeding is self-contained.
 **Plans:** 4/4 plans complete
 Plans:
 **Wave 1**
 - [x] 18-01-PLAN.md — TDD: getHouseholdTimezone(db) accessor + isValidIanaTimezone (D-05/D-06)
 **Wave 2** *(blocked on Wave 1 completion)*
 - [x] 18-02-PLAN.md — TDD: admin GET/PUT/seed timezone endpoints on adminRouter, requireAdmin + IANA validation + no-overwrite seed (D-01/D-02/D-03/D-04)
 - [x] 18-03-PLAN.md — TDD: route all-day reminder TZ at reminderScheduler:247 + outboxWorker:501,607 through the accessor (D-05/D-06/D-07)
 **Wave 3** *(blocked on Wave 2 completion)*
 - [x] 18-04-PLAN.md — PWA Timezone section in /admin Settings (searchable IANA picker + detected-zone seed) + client fns (D-02/D-04)
 **UI hint**: yes
 ### Phase 19: Local Auth (No-OIDC Mode)
 **Goal:** Let an operator run FamilySync entirely on **local DB users with no OIDC** — username/password accounts and a local login flow that coexists with the Authelia OIDC path — and **optionally wire OIDC in later** by claiming/linking an existing local user to an OIDC identity. Removes the hard dependency on a deployed Authelia for small/solo self-hosters.
 **Mode:** standard
 **Depends on:** Phase 12 (Initial Setup Wizard) — builds directly on the pre-OIDC **local-user foundation** introduced there: nullable `users.oidc_iss`/`oidc_sub` + the claimed/pending marker, and the first-login-claims merge. Phase 19 generalizes that single bootstrap local user into a full local-account model + login.
 **Requirements**: AUTH-LOCAL-01..AUTH-LOCAL-20 (derived during planning 2026-06-17) — local_credentials schema (01), scrypt hash/verify (02), login route (03), localAuthMiddleware (04), auth-mode endpoint (05), logout (06), admin create-member (07), admin reset (08), self-change (09), OIDC-link (10), break-glass CLI (11), LoginPage (12), admin UI (13), settings UI (14), routing gate (15), dev-bypass/harness rework (16), hasLocalCredential (17), de-Authelia copy (18), rate-limit/lockout (19), auth unit tests (20). Plus `LOCAL_SESSION_SECRET` env + boot assertion (D-05).
 **Plans:** 5/5 plans complete
 **Provenance:** Deferred from the Phase 12 discussion (2026-06-15) — see `.planning/phases/12-initial-setup-wizard/12-CONTEXT.md` §Deferred Ideas. The operator runs FamilySync this way themselves and wants no-OIDC operation as a first-class mode.
 **Open questions for discuss/spec:**
 - Password hashing/storage choice (e.g. argon2id/bcrypt) and how it sits alongside the env-only secret kernel from Phase 12.
 - How local login coexists with `oidcAuthMiddleware` ordering in `apps/api/src/index.ts` (route-level auth strategy selection vs. a mode flag in `app_config`).
 - The OIDC-link flow: claiming an existing local user into an `oidc_iss+oidc_sub` identity without violating the D-10 "identity is OIDC, never email" rule.
 - Whether "local mode vs OIDC mode" is a deploy-time switch or both can be live simultaneously.
 Plans:
 **Wave 1**
 - [x] 19-01-PLAN.md — Foundation (TDD): local_credentials schema + 0003 migration, scrypt hash/verify, local-session JWT helpers, LOCAL_SESSION_SECRET boot guard + generate-secrets, .dockerignore scripts exclusion (AUTH-LOCAL-01/02)
 **Wave 2** *(blocked on Wave 1)*
 - [x] 19-02-PLAN.md — Backend account mgmt (TDD): admin create/reset member, self-change password, hasLocalCredential, linkOidcToUser helper + /api/me/link-oidc (AUTH-LOCAL-07/08/09/10/17)
 **Wave 3** *(blocked on Wave 2)*
 - [x] 19-03-PLAN.md — Middleware + routes + wiring (TDD): localAuthMiddleware, /api/auth/mode, login (rate-limit/lockout) + logout, index.ts mount + OIDC-guard skip + /callback link branch, de-Authelia comments (AUTH-LOCAL-03/04/05/06/18/19/20)
 **Wave 4** *(blocked on Wave 3; 04 + 05 parallel)*
 - [x] 19-04-PLAN.md — PWA: LoginPage + BrandSlot + App.tsx gate + client.ts + AdminPage + SettingsSheet (AUTH-LOCAL-12/13/14/15)
 - [x] 19-05-PLAN.md — Dev-bypass Option C + break-glass CLI + harness/CI rework + login.spec.ts (AUTH-LOCAL-11/16)
 ### Phase 20: Admin Member Editor & Form Declutter
 **Goal:** Replace the per-member-row action buttons (Rotate/Add credential + Reset password) in the admin Members panel with a single edit affordance — clicking a member's name or an edit button opens a member-detail editor where an admin modifies all of that member's details in one place: display name, local-login password, and the Fastmail/CalDAV app password (calendar credential) — using clear, non-jargon labels that retire the confusing "Rotate" term. Also collapse the "Add member" section so its input fields are hidden behind a single "Add member" trigger by default, decluttering the panel. Client-side AdminPage + CredentialSheet rework over the existing `/api/admin` endpoints; no new auth/authorization boundary (seeded by the gripe that "Rotate" for the app password is not intuitive).
 **Requirements**: TBD (refine in /gsd-discuss-phase 20 — open scope: which fields count as "all" (color swatch? admin toggle? OIDC link?), whether to keep any standalone reset-password flow, and the exact edit affordance — clickable name vs. row edit button)
 **Depends on:** Phase 19
 **Plans:** 3/3 plans complete
 Plans:
 **Wave 1**
 - [x] 20-01-PLAN.md — Server: PATCH /api/admin/members/:id (displayName + is_admin) with last-admin demotion guard (TDD) + isAdmin in GET /members
 - [x] 20-02-PLAN.md — PWA API client: AdminMember.isAdmin field + updateMemberProfile fetcher (last-admin sentinel)
 **Wave 2** *(blocked on Wave 1 completion)*
 - [x] 20-03-PLAN.md — PWA: unified MemberEditorSheet (edit/create, per-section saves) + decluttered tappable Members panel; retire Rotate/Reset-password buttons
@@ -2,35 +2,36 @@
 gsd_state_version: 1.0
 milestone: v1.1
 milestone_name: Operability & Polish
-current_phase: 999.1
+current_phase: null
-current_phase_name: BACKLOG
+status: Awaiting next milestone
-status: "Phase 20 shipped — PR #25"
+stopped_at: v1.1 milestone shipped & archived
-stopped_at: Phase 20 UI-SPEC approved
+last_updated: "2026-06-19T01:58:30.566Z"
 last_updated: "2026-06-19T00:26:05.257Z"
 last_activity: 2026-06-18
 last_activity_desc: Milestone v1.1 completed and archived
 progress:
-  total_phases: 27
+  total_phases: 20
-  completed_phases: 13
+  completed_phases: 20
-  total_plans: 58
+  total_plans: 57
  completed_plans: 57
-  percent: 48
+  percent: 100
 current_phase_name: Awaiting next milestone
 ---
 # Project State
 ## Project Reference
-See: .planning/PROJECT.md (updated 2026-06-16)
+See: .planning/PROJECT.md (updated 2026-06-18 after v1.1 milestone)
 **Core value:** One color-coded family calendar (shared + personal) and shared lists from a single low-friction PWA — cross-ecosystem, no app store
-**Current focus:** Phase 20 — admin-member-editor-form-declutter
+**Current focus:** Planning next milestone — run `/gsd-new-milestone`
 ## Current Position
-Phase: 999.1 — Treat Fastmail as one calendar provider; framework supports adding more providers (BACKLOG)
+Phase: Milestone v1.1 complete
-Plan: Not started
+Plan: —
-Status: Phase 20 shipped — PR #25
+Status: Awaiting next milestone
-Last activity: 2026-06-18
+Last activity: 2026-06-19 — Milestone v1.1 completed and archived
 ### ✅ Resolved Checkpoint — Phase 15 Plan 15-03 Task 2 (human-action)
@@ -256,6 +257,8 @@ Recent decisions affecting current work:
 | 260613-dmw | Exclude `.gitea/**` from the CI `changes` `code` paths-filter so workflow-only PRs skip the heavy api/harness jobs (treated like docs) while fast-checks + gate still run. Single `- '!.gitea/**'` negation appended after the yml/yaml globs (index 11 vs 5). Rides along on the Phase 16 branch / PR #15. | 2026-06-13 | 2d329a9 |  | [260613-dmw-exclude-gitea-workflow-config-changes-fr](./quick/260613-dmw-exclude-gitea-workflow-config-changes-fr/) |
 | 260613-fp9 | `.gitea`/`.planning`-only pushes to main no longer trigger the Docker publish — added `paths-ignore: ['.gitea/**', '.planning/**']` under `on.push` in `.gitea/workflows/publish.yml` (skips only when EVERY changed file matches; mixed code+docs pushes still publish). `.dockerignore` already excludes `.planning` so the image is byte-identical. Done in isolated worktree (phase-10 agent held main tree). | 2026-06-13 | cd5a88c | | [260613-fp9-gitea-and-planning-pushes-should-not-tri](./quick/260613-fp9-gitea-and-planning-pushes-should-not-tri/) |
 | 260613-ndv | Isolate local apps/api integration tests to a dedicated `familysync_test` DB so test runs stop polluting the dev `familysync` DB. New CI-gated vitest globalSetup root-provisions (CREATE DATABASE + GRANT) + migrates + truncate-resets `familysync_test` each run; `vitest.config.ts` forces `DB_NAME=familysync_test` for local workers (no-op under CI, so CI's `familysync` service DB + db:migrate are untouched). Verified: dev `familysync` users stays 3 across a run, `familysync_test` resets (186→93, not doubled), 244/244 tests pass (flaky list_shares timeout gone), typecheck 0. Branch off main. | 2026-06-13 | 07d5161 | Verified | [260613-ndv-wire-apps-api-integration-tests-to-a-ded](./quick/260613-ndv-wire-apps-api-integration-tests-to-a-ded/) |
 | 260618-smr | Remove unused Redis service and all references — Redis confirmed unused at runtime (no ioredis/redis client import, no `REDIS_*` env, not a dependency in any package.json). Dropped the `redis` service from both compose files and cleaned all references in CLAUDE.md, README.md, and docs/* + e2e config. Kept the in-memory-vs-Redis design-rationale comments (D-12/D-18) in listEmitter/reminderScheduler/linkNonceStore/localAuth. `docker compose config` parses clean (0 redis); `format:check` green. Branch off main. | 2026-06-18 | 0b42666 | Verified | [260618-smr-remove-unused-redis-service-and-referenc](./quick/260618-smr-remove-unused-redis-service-and-referenc/) |
 | 260618-tg2 | Persistent CI dependency caches — point all 4 CI `pnpm install` steps at a host-mounted `/pnpm-store` (`--store-dir /pnpm-store --prefer-offline`) and persist Playwright browsers via `PLAYWRIGHT_BROWSERS_PATH=/ms-playwright` on the harness job; added BuildKit `--mount=type=cache` to all 3 Dockerfile install stages + `DOCKER_BUILDKIT=1` on the publish build. Avoids `actions/cache` (D-PROBE-04 timeout). In-repo only — requires act_runner `config.yaml` `container.options` host mounts (manual host change). Verdaccio deferred. Branch off main. | 2026-06-18 | 6e93e24 | Verified | [260618-tg2-persistent-ci-dependency-caches-pnpm-sto](./quick/260618-tg2-persistent-ci-dependency-caches-pnpm-sto/) |
 ## Deferred Items
@@ -281,5 +284,4 @@ Resume file: .planning/phases/20-admin-member-editor-form-declutter/20-UI-SPEC.m
 ## Operator Next Steps
- **Phase 8 is complete.** CI pipeline is fully operational on the self-hosted Gitea runner.
+- Start the next milestone with /gsd-new-milestone
 - Next: `/gsd-plan-phase 9` (Faster Write-Back — fully independent, lowest risk) or `/gsd-plan-phase 10` (Admin Role & Settings — carries the v1.1 DB migration that Phases 11 & 12 depend on). These can run in parallel once planned.
@@ -1,3 +1,12 @@
 # Requirements Archive: v1.1 Operability & Polish
 **Archived:** 2026-06-19
 **Status:** SHIPPED
 For current requirements, see `.planning/REQUIREMENTS.md`.
 ---
 # Requirements: FamilySync — v1.1 "Operability & Polish"
 **Defined:** 2026-06-10
@@ -0,0 +1,772 @@
 # Roadmap: FamilySync
 ## Milestones
 - ✅ **v1.0 MVP** — Phases 1–6 (shipped 2026-06-10) — see [`milestones/v1.0-ROADMAP.md`](milestones/v1.0-ROADMAP.md)
 - ✅ **v1.1 Operability & Polish** — Phases 7–20 (shipped 2026-06-18) — mobile test harness, Gitea CI (runs the harness), faster write-back, in-app admin, per-event reminders, guided setup, real lint gate, desktop e2e, doc-only CI skip + markdown lint, CI dependency audit + security checks + image hygiene, UI optimization & polish, auto timezone detection, local auth (no-OIDC mode), admin member editor & declutter
 ## Phases
 <details>
 <summary>✅ v1.0 MVP (Phases 1–6) — SHIPPED 2026-06-10</summary>
 - [x] Phase 1: Foundation + Broker Spike (4/4 plans) — completed 2026-06-04
 - [x] Phase 2: Calendar Display (5/5 plans) — completed 2026-06-05
 - [x] Phase 3: Event Write-Back + PWA Install (12/12 plans) — completed 2026-06-07
 - [x] Phase 4: Shared Lists + Live Sync (7/7 plans) — completed 2026-06-09
 - [x] Phase 5: Web Push Notifications (8/8 plans) — completed 2026-06-10
 - [x] Phase 6: UX Polish (6/6 plans) — completed 2026-06-10
 Full phase detail archived in [`milestones/v1.0-ROADMAP.md`](milestones/v1.0-ROADMAP.md).
 </details>
 ### ✅ v1.1 Operability & Polish (Phases 7–20) — SHIPPED 2026-06-18
 Make FamilySync configurable, administrable, and maintainable for real multi-member use — without hand-editing env files or the database. The new critical path runs **mobile test harness → Gitea CI** (CI consumes the harness specs for UI regression), and the **admin role → reminders / setup wizard** chain (a single `/api/admin` + `/api/setup` route surface carrying the v1.1 DB migration). Faster write-back is a fully independent track.
 - [x] **Phase 7: Mobile Test Harness** - Mobile-emulated, authenticated PWA browser harness so the assistant (and CI) can catch mobile-only defects (completed 2026-06-11)
 - [x] **Phase 8: Gitea CI** - Full regression on PR to main (lint/typecheck/unit/API-integration vs a MariaDB service container **+ the Phase 7 mobile harness as a UI-regression step against a CI-hosted dev stack**) + Docker image publish on merge (completed 2026-06-11)
 - [x] **Phase 9: Faster Write-Back** - Event-driven outbox drain so edits land in ~1-2s instead of ~15s, preserving every outbox durability guarantee (completed 2026-06-12)
 - [x] **Phase 10: Admin Role & Settings** - DB foundation (is_admin / reminder_lead / app_config) + role-gated admin UI to rotate app passwords and designate the shared calendar (completed 2026-06-13)
 - [x] **Phase 11: Per-Event Reminders** - Reminder selector on the event form (incl. "None") serialized as VALARM, with a variable-lead scheduler that honors each event's choice (completed 2026-06-14)
 - [x] **Phase 12: Initial Setup Wizard** - First-run validated bootstrap of env/VAPID/DB/OIDC + first app password, reusing the admin route surface (completed 2026-06-16)
 - [x] **Phase 13: Real Lint Gate (ESLint)** - Wire ESLint flat config (typescript-eslint + React) across both apps so the Phase 8 CI lint slot actually fails on violations instead of no-op'ing (completed 2026-06-12)
 - [x] **Phase 14: Desktop E2E Coverage** - Add a Desktop Chrome Playwright profile + make the mobile-authored specs desktop-safe so the Phase 8 regression gate validates desktop, not just mobile (completed 2026-06-12)
 - [x] **Phase 15: Doc-Only CI Skip + Markdown Lint** - Aggregate-gate the slow api/harness CI jobs so doc-only PRs to main merge without running them (no branch-protection deadlock), and add markdownlint to `fast-checks` so docs get a fast format+lint gate (promoted from backlog 999.17) (completed 2026-06-12)
 - [x] **Phase 16: CI Dependency Audit, Security Checks & Image Hygiene** - Extend Gitea CI with outdated-dependency reporting + vulnerability audit + a baseline of additional security checks, and enforce the dev/prod image boundary so no dev-bypass, secret, or family data ships in published images (absorbs backlog 999.17); independent of the admin chain (completed 2026-06-13)
 - [x] **Phase 17: UI Optimization & Polish** - Phone-layout polish + branding + theme groundwork: fix the long-standing phone-layout overlap where the fixed BottomTabBar covers the New Event FAB and the colour legend (+ small-viewport sweep), finish the branding assets (real FamilySync logo into the BrandSlot seam + a complete favicon/PWA-icon set replacing the placeholder stubs), and restructure tokens.css into a themeable token layer (light-only groundwork for future dark mode). Shipped dark theme → backlog 999.20; broader styling refresh → backlog 999.21 (future milestone) (completed 2026-06-18)
 ## Phase Details
 > v1.0 phase detail (Phases 1–6) is archived in [`milestones/v1.0-ROADMAP.md`](milestones/v1.0-ROADMAP.md).
 ### Phase 7: Mobile Test Harness
 **Goal**: The assistant can drive the PWA in a mobile-emulated, authenticated browser context against the host-side dev stack, so mobile-only layout and flow defects can be caught automatically instead of only by the operator on real devices. This harness is also the artifact Phase 8 (CI) runs for UI regression.
 **Mode:** standard
 **Depends on**: Nothing (fully independent; goes first. One new dev dependency `@playwright/test` in `apps/pwa`; no backend changes).
 **Requirements**: TEST-01, TEST-02
 **Success Criteria** (what must be TRUE):
  1. An automated run can load the PWA in a mobile-emulated viewport (device profile + mobile UA + touch) and assert on responsive layout / tap targets.
  2. The automated run reaches the authenticated PWA via the existing `DEV_AUTH_BYPASS` on the host-side dev stack — no manual login and no Authelia/OIDC mocking.
  3. The harness runs repeatably day-over-day without re-capturing any session state (no stale storage-state failures).
  4. The harness specs are structured so they can run headlessly in CI (Phase 8) against a dev stack the runner brings up — no dependence on a developer's already-running host stack.
 **Pitfalls this phase owns** (from PITFALLS.md):
  - **No stale storage-state** (Pitfall 14): use `DEV_AUTH_BYPASS=true` for the automated harness rather than a checked-in storage-state.json with an expiring session cookie; decide the auth strategy before the first test.
  - **Service worker block** (Pitfall 15): set `serviceWorkers: 'block'` (or explicitly unregister) in the context so a previous run's SW does not intercept requests / return stale cached responses; verify no SW-sourced responses in the trace.
  - Hard constraints: targets the dev build via `DEV_AUTH_BYPASS` (DEV_AUTH_BYPASS user 1 has no CalDAV credential/calendars — verify layout/flows, not live event-create); real prod-service-worker / iOS-Safari-standalone mobile testing stays a human/device gate (out of scope).
 **Plans**: 4 plans (3 waves)Plans:
 **Wave 1**
 - [x] 07-01-PLAN.md — Harness foundation: @playwright/test + WebKit/Chromium browsers, playwright.config.ts (iPhone/WebKit + Pixel/Chromium matrix, serviceWorkers block, env baseURL, vite webServer), vitest exclude, scripts (Wave 1)
 **Wave 2** *(blocked on Wave 1 completion)*
 - [x] 07-02-PLAN.md — global-setup.ts: /health readiness poll + deterministic mysql2 reset-and-seed (calendar id 10 INSERT IGNORE guard, list + items) + e2e README/guardrails (Wave 2)
 **Wave 3** *(blocked on Wave 2 completion)*
 - [x] 07-03-PLAN.md — layout.spec.ts: tap targets >=44px, no overflow, in-viewport, accessible names (UI-SPEC Rules 1-4) + harness self-validation injected-defect proofs (Wave 3)
 - [x] 07-04-PLAN.md — calendar.spec.ts + lists.spec.ts: populated/empty/error states (Rules 4/5) + DEV_AUTH_BYPASS auth-reached + no-SW-controller precondition (Wave 3)
 **UI hint**: yes
 ### Phase 8: Gitea CI
 **Goal**: Every PR to `main` runs a full regression that gates the merge — lint, typecheck, unit, API-integration against a MariaDB service container, **and the Phase 7 mobile Playwright harness as a UI-regression step against a CI-hosted dev stack** — and a merge to `main` builds and publishes the API Docker image, all on the existing self-hosted Gitea Actions runner.
 **Mode:** standard
 **Depends on**: Phase 7 (the PR regression runs the Phase 7 mobile harness specs as its UI-regression step; without the harness there is nothing to run). No other code dependencies. Start with a runner-probe step.
 **Requirements**: CI-01, CI-02
 **Success Criteria** (what must be TRUE):
  1. Opening or updating a PR targeting `main` triggers a workflow that runs lint, typecheck (both apps), unit tests, and API integration tests against a MariaDB service container — and a failing run blocks the merge.
  2. The API integration tests connect to the service-container MariaDB (DB_HOST=127.0.0.1, service creds) and pass reliably on a cold first run, not only on re-run.
  3. The same PR workflow brings up the dev stack inside the runner — the API dev server, the PWA dev server, and the MariaDB service container, with `DEV_AUTH_BYPASS=true` — and runs the Phase 7 mobile Playwright harness specs headlessly against that authed PWA; a harness failure blocks the merge.
  4. The harness step waits for both the API and PWA dev servers to be ready (readiness probe / poll) before launching Playwright, so it does not flake on startup races.
  5. On merge to `main`, the API Docker image is built and pushed to the Gitea container registry under a sensible tag.
  6. Registry credentials never appear in plaintext in the CI logs.
 **Pitfalls this phase owns** (from PITFALLS.md):
  - **Runner-probe first** (Pitfall 12): the first workflow only probes `node --version` / `pnpm --version` / Docker access on the `self-hosted` runner before any test or build steps are designed; pin Node 22 explicitly, do not assume `actions/setup-node` works as on GitHub.
  - **MariaDB readiness wait** (Pitfall 11): add an explicit readiness loop (e.g. `healthcheck.sh --connect --innodb_initialized`, NOT `mysqladmin ping` which is removed in MariaDB 11) before any `drizzle-kit migrate` / integration test step; healthy ≠ accepting connections.
  - **Dev-stack readiness races (NEW for the harness step):** running the PWA and API dev servers *inside* CI adds startup/readiness races on top of the MariaDB-11 readiness race. The harness step must wait for **both** the API and PWA dev servers to be accepting connections (poll their URLs / health endpoints) before Playwright launches — do not race the browser against a not-yet-listening server. Run with `DEV_AUTH_BYPASS=true` so the harness reaches the authed PWA exactly as in Phase 7.
  - **--password-stdin** (Pitfall 13): `docker login` via `--password-stdin` with the token piped from a registered Gitea secret (PAT with `write:package`); never `-p $TOKEN` on the command line.
  - Hard constraints: API integration tests need a real MariaDB and live in `apps/api/tests/` (never `src/`); cache the pnpm store; Drizzle generate+migrate to set up the CI DB schema; the harness step reuses the Phase 7 specs unchanged (CI owns only the stack bring-up + readiness wait, not the spec content).
 **Plans**: 4 plans (4 waves)Plans:
 **Wave 1**
 - [x] 08-01-PLAN.md — Runner probe + operator runner/PAT registration (W0; answers the Docker-vs-host fork)
 **Wave 2** *(blocked on Wave 1 completion)*
 - [x] 08-02-PLAN.md — ci.yml: fast-checks (lint/typecheck/PWA unit) + API job (MariaDB service + migrate + DB-backed tests)
 **Wave 3** *(blocked on Wave 2 completion)*
 - [x] 08-03-PLAN.md — ci.yml: harness job (dev-stack bring-up + readiness waits + Phase 7 Playwright specs, both profiles)
 **Wave 4** *(blocked on Wave 3 completion)*
 - [x] 08-04-PLAN.md — ci.yml: publish job (build production image + push :latest + :v1.1-<sha> via --password-stdin)
 **UI hint**: yes
 ### Phase 9: Faster Write-Back
 **Goal**: A created, edited, or deleted event reaches Fastmail within ~1-2 seconds (event-driven outbox drain) instead of waiting up to ~15s for the next interval tick — with every existing durability guarantee intact.
 **Mode:** standard
 **Depends on**: Nothing (fully independent; the only new artifact is a zero-dependency in-process EventEmitter, `lib/outboxTrigger.ts`). Can run in parallel with any other v1.1 track.
 **Requirements**: CAL-15
 **Success Criteria** (what must be TRUE):
  1. After creating/editing/deleting an event, the change lands in Fastmail in ~1-2s in the common case (drain is signalled on enqueue, not waited-for on the interval) — observable as the change appearing in the Fastmail native app well before the old ~15s window.
  2. The route handler still returns an optimistic 202 immediately and never makes a CalDAV call inline — the event-driven signal is fire-and-forget.
  3. Edit-as-move still writes the new event before deleting the old one (create-before-delete ordering preserved); no event is ever lost when a move drains under rapid enqueues.
  4. No duplicate CalDAV PUTs occur for the same outbox row when the signal and the 15s fallback interval overlap (exactly-once per uid preserved).
  5. The 15s `setInterval` fallback still runs and recovers any rows missed by the signal path (startup catch-up, transient errors).
 **Pitfalls this phase owns** (from PITFALLS.md):
  - **No double-drain** (Pitfall 5): the trigger must set a `drainRequested` flag funnelled through the single setInterval-controlled path / the existing `isDraining` guard — never call `runOutboxDrain()` directly from the signal in a way that bypasses the guard or escapes the error-caught wrapper.
  - **Create-before-delete under concurrent enqueues** (Pitfall 6): enqueue CREATE before DELETE; do not fire the signal between the two inserts of a move (publish after both inserts / after the transaction commits).
  - Hard constraints: `setInterval` only (no node-cron); single-process by design — **no Redis** for the drain (Redis stays for list SSE); all outbox guarantees (fresh-etag-before-PUT, 412 conflict flow, per-uid exactly-once) unchanged.
 **Plans**: 2 plans (2 waves)
 Plans:
 **Wave 1**
 - [x] 09-01-PLAN.md — TDD: outboxTrigger.ts (zero-dep EventEmitter signal) + scheduleOutboxDrain wrapper / drainRequested trailing-re-drain loop + initOutboxTrigger in outboxWorker.ts; trigger-wiring tests (SC-1, SC-4/D-07, D-05) (Wave 1)
 **Wave 2** *(blocked on Wave 1 completion)*
 - [x] 09-02-PLAN.md — Four post-commit signalOutboxDrain() publish sites in events.ts (create / edit-as-move-after-transaction / same-cal update / delete) + initOutboxTrigger() startup wiring under isMainModule() in index.ts (Wave 2)
 ### Phase 10: Admin Role & Settings
 **Goal**: An admin can manage household configuration that previously required manual DB writes — rotating a member's Fastmail app password and designating the shared family calendar — from a role-gated in-app Settings section, on top of the v1.1 DB foundation this phase introduces.
 **Mode:** standard
 **Depends on**: Nothing required upstream; this phase **carries the v1.1 DB migration** (users.is_admin, calendar_events.reminder_lead_minutes, app_config table) that Phases 11 and 12 build on. It is the head of the admin chain (10 → 11, 10 → 12).
 **Requirements**: ADMIN-01, ADMIN-02, ADMIN-03
 **Success Criteria** (what must be TRUE):
  1. An admin sees an Admin section in Settings and can list household members with their credential status; a non-admin member never sees it and cannot invoke any `/api/admin/*` route (gets 403).
  2. An admin can enter or rotate a member's Fastmail app password; it is validated against CalDAV (PROPFIND) before saving and stored encrypted — and the password is never displayed, echoed in a response, or logged.
  3. An admin can pick which synced calendar is the shared family calendar from a list, and the `calendars.is_shared` flag updates accordingly (replacing the manual `UPDATE calendars SET is_shared=1` step).
  4. The role check is role-agnostic and member-count-agnostic: it gates on `users.is_admin`, so more admins can be added later without reworking the guard.
  5. The DB migration (is_admin, reminder_lead_minutes, app_config) is applied via generate+migrate and is in place for downstream phases (reminder_lead_minutes for Phase 11, app_config.setup_complete for Phase 12).
 **Pitfalls this phase owns** (from PITFALLS.md):
  - **Admin role check inside the sub-router** (Pitfall 9): apply `requireAdmin` with `.use('*', ...)` inside `adminRouter`, not only at the parent mount; integration test must assert 403 for a non-admin authenticated user.
  - **App password never logged/echoed** (Pitfall 7): custom zod-validator `hook` returns a generic 400 (no Zod `received`/`value` field); no `console.log` of request bodies in `routes/admin*`.
  - Hard constraints: Drizzle **generate+migrate, never push** (false destructive diff on populated MariaDB); reuse `broker/crypto.ts` `encryptPassword` (no changes to crypto); `/api/admin/credentials` and `/api/admin/calendars/:id/shared` are the single shared surface — do NOT duplicate them into `/api/setup/*` in Phase 12.
 **Folded-in scope** (from backlog 999.5, self-service member onboarding): the credential surface this phase builds is the same one a member needs on first login. Expose a `needsProviderSetup` signal (member has no `member_credentials` row) and let a member enter/validate (CalDAV PROPFIND) + encrypt their **own** Fastmail app password — the self-service counterpart of the admin-managed flow, sharing the validation/encryption/initial-sync path. Non-technical-friendly instructions (link to Fastmail's app-password page, required Calendars/CalDAV scope) are the hard UX constraint. Member-scoped: a member can only set their own credential; never log/echo the password.
 **Plans**: 4 plans (4 waves)Plans:
 **Wave 1**
 - [x] 10-01-PLAN.md — v1.1 DB foundation migration (is_admin, provider_type+unique, reminder_lead_minutes, app_config) + dev-bypass admin seed
 **Wave 2** *(blocked on Wave 1 completion)*
 - [x] 10-02-PLAN.md — requireAdmin guard + first-login-wins bootstrap + /api/me isAdmin/needsProviderSetup (TDD)
 **Wave 3** *(blocked on Wave 2 completion)*
 - [x] 10-03-PLAN.md — adminRouter (members/credentials/calendars/shared) + member self-service credential, validate→encrypt→sync (TDD)
 **Wave 4** *(blocked on Wave 3 completion)*
 - [x] 10-04-PLAN.md — PWA /admin route + nav gating + CredentialSheet + SetupBanner (playwright-cli verified)
 **UI hint**: yes
 ### Phase 11: Per-Event Reminders
 **Goal**: A user can choose a reminder lead time per event (None / 5m / 10m / 15m / 30m / 1h / 2h / 1d / 2d, default None), serialized as a VALARM on the event, and the push scheduler fires at that exact lead — firing nothing when there is no alarm and never stripping reminders set in other clients.
 **Mode:** standard
 **Depends on**: Phase 10 (the `calendar_events.reminder_lead_minutes` column from the v1.1 migration is the scheduler's ground truth). Independent of Phases 7/8/9/12.
 **Requirements**: CAL-13, CAL-14, NOTIF-04, NOTIF-05, NOTIF-06
 **Success Criteria** (what must be TRUE):
  1. When creating or editing a timed event, the user can pick a reminder lead from the preset list (None default); the choice round-trips to Fastmail as a VALARM and is visible/honored on re-open.
  2. Editing an event that already has a reminder set in another client (Fastmail / Apple Calendar) preserves that VALARM — it is never silently dropped on round-trip.
  3. A reminder push fires at the event's chosen lead time (e.g. T-30 for a 30-minute lead), not a hardcoded 15-minute lead.
  4. An event with no reminder set produces no reminder push (no default 15-minute fire).
  5. An all-day event's reminder fires at a sensible local time (9 AM on the alert day), not midnight; the reminder selector is disabled/hidden for all-day events in the UI; and reminder delivery stays exactly-once across catch-up scans and rescheduled events.
 **Pitfalls this phase owns** (from PITFALLS.md):
  - **Preserve-on-edit** (Pitfall 1): the update path extracts and preserves existing VALARM sub-components from `rawVevent` (mirroring the WR-01 RRULE-preserve pattern) — never rebuild-from-scratch and silently strip; `outboxPayloadSchema` distinguishes "no change" from explicit "no reminder".
  - **No TRIGGER VALUE=TEXT** (Pitfall 2): build the trigger with `ICAL.Duration.fromSeconds(-n*60)`, not a bare string; unit-test that the ICS emits a DURATION trigger with no `VALUE=TEXT`.
  - **All-day 9AM semantics** (Pitfall 3): guard `buildVeventString` (`if (!allDay && reminderMinutes > 0)`), disable the selector when allDay, keep the scheduler's all-day handling at 9 AM local.
  - **uid:dtstartMs dedup** (Pitfall 4): change the scheduler dedup key from bare `uid` to compound `uid:dtstartMs` and widen the scan to a variable per-event window so long leads fire and rescheduled events re-fire; keep `eventFieldsSchema` and `outboxPayloadSchema` in sync (IN-03).
  - Hard constraints: `setInterval` only; scheduler reads `reminder_lead_minutes` from the DB (ground truth), not the outbox payload; drop the `isShared`-only reminder restriction (a user who set an alarm wants it regardless of calendar).
 **Plans**: 4 plans (3 waves)
 Plans:
 **Wave 1**
 - [x] 11-01-PLAN.md — VALARM builders + classifier + extractor + computeAlertInstantUtc (vevent.ts, TDD)
 **Wave 2** *(blocked on Wave 1 completion)*
 - [x] 11-02-PLAN.md — Variable-lead scheduler: uid:dtstartMs dedup, drop isShared, all-day 9 AM, humanized body (TDD)
 - [x] 11-03-PLAN.md — Backend plumbing: schema field, outbox preserve-on-edit, sync upsert, occurrence surfacing
 **Wave 3** *(blocked on Wave 2 completion)*
 - [x] 11-04-PLAN.md — EventForm reminder picker (allDay swap, edit pre-population) + client types + Playwright smoke
 **UI hint**: yes
 ### Phase 12: Initial Setup Wizard
 **Goal**: On first run (no admin/credentials configured), the operator is guided through a validated, step-by-step wizard to bootstrap the app — env presence, generated secrets to copy, DB/OIDC/VAPID/app-password validation — instead of hand-editing `.env` / `docker-compose.yml`; once complete, the setup endpoints lock.
 **Mode:** standard
 **Depends on**: Phase 10 (reuses the admin role + `/api/admin/credentials` and `/api/admin/calendars/:id/shared` routes; the wizard is the second frontend consumer of that surface, and `app_config` from the Phase 10 migration holds `setup_complete`). Goes last. Independent of Phases 7/8/9/11.
 **Requirements**: SETUP-01, SETUP-02, SETUP-03, SETUP-04
 **Success Criteria** (what must be TRUE):
  1. On a fresh install with nothing configured, the operator reaches a setup wizard (via `GET /api/setup/status` mounted before the OIDC guard) and walks through bootstrap steps instead of editing files by hand.
  2. Each input is validated before the step can complete: DB connects, VAPID private key decodes to exactly 32 bytes and pairs with the public key, OIDC discovery resolves, and the Fastmail app password reaches CalDAV (PROPFIND).
  3. Generated secrets (session secret, encryption key, VAPID keypair) are displayed for the operator to copy into env; they are never written to the DB or returned in a way that persists, and `APP_PASSWORD_ENCRYPTION_KEY`/`VAPID_PRIVATE_KEY` never enter the DB at all.
  4. After completion, the wizard-completing user is promoted to admin (`is_admin`), `app_config.setup_complete` is set, and any further call to a setup endpoint returns 423 Locked.
  5. The 423 guard is enforced on every invocation (checked against member-credentials + VAPID env present), not only at startup.
 **Pitfalls this phase owns** (from PITFALLS.md):
  - **Guard on every invocation** (Pitfall 8): the "already set up" guard returns 423 from all setup routes once configured — implement and test the guard before the happy path; a second POST after completion must return 423, not 200.
  - **Secrets stay in env, never in DB** (Pitfalls 8 & 10): the wizard validates secrets by performing a test operation (test encrypt/decrypt, structural VAPID check), never by accepting/storing the key value; no DB column for `vapid_private_key` or `app_password_encryption_key`; never log/echo the app password.
  - Hard constraints: `GET /api/setup/status` mounts **before** the OIDC guard (like `/health`); do NOT create `/api/setup/credentials` — reuse the Phase 10 admin routes; Drizzle generate+migrate (any `app_config` seeding via migration).
 **Plans**: 7 plans in 4 waves (4 original + 3 gap-closure for 12-UAT.md gaps 1-6)
 Plans:
 **Wave 1**
 - [x] 12-01-PLAN.md — Schema migration (nullable OIDC + claimed) + generate-secrets helper (SETUP-03) + Wave-0 scaffolds
 **Wave 2** *(blocked on Wave 1 completion)*
 - [x] 12-02-PLAN.md — Pre-auth /api/setup/* router + isSetupLocked 423 guard + index mount + OIDC boot fallback (SETUP-01/02/04)
 - [x] 12-03-PLAN.md — First-login-claims rework in upsertUser (D-08, SETUP-01)
 **Wave 3** *(blocked on Wave 2 completion)*
 - [x] 12-04-PLAN.md — PWA SetupPage wizard + App.tsx gate + UI-SPEC revision (SETUP-01/02)
 **Wave 4 — Gap closure** *(UAT 12-UAT.md gaps 1-6; 06+07 parallel, 05 blocked on 06)*
 - [x] 12-06-PLAN.md — Backend: validate/vapid asserts wizard key == env VAPID_PUBLIC_KEY (gap 2) + status exposes non-secret DB name (gap 3) (SETUP-02)
 - [x] 12-07-PLAN.md — App.tsx: reverse-gate /setup post-completion (gap 5) + reconcile ['me'] so calendar banner clears after wizard (gap 6) (SETUP-01/04)
 - [x] 12-05-PLAN.md — SetupPage: drop DB-vs-env aside (gap 1) + read-only DB-name field (gap 3) + persist fields across Back (gap 4) (SETUP-01) — depends on 12-06
 **UI hint**: yes
 ### Phase 13: Real Lint Gate (ESLint)
 **Goal**: The CI lint gate actually fails on lint violations. A real ESLint flat config (`eslint.config.js`, `typescript-eslint`; React + react-hooks plugins for `apps/pwa`) plus a package-level `lint` script in `apps/api` and `apps/pwa` makes the existing root `pnpm -r --if-present lint` run a real linter, replacing the hollow no-op gate that exits 0 because no linter exists.
 **Mode:** standard
 **Depends on**: Phase 8 (the CI `fast-checks` job already runs `pnpm lint`; this fills the slot Phase 8 shipped wired to auto-activate once a package `lint` script lands). Independent of all other phases.
 **Requirements**: TBD (promoted from backlog 999.16)
 **Success Criteria** (what must be TRUE):
  1. `pnpm lint` runs ESLint across both `apps/api` and `apps/pwa` and exits non-zero on an introduced violation (verified by a deliberate test violation), where today it exits 0 with no linter present.
  2. The CI `fast-checks` lint step blocks a PR to main on lint violations — the gate can now fail.
  3. The first real run's existing violations are resolved (fix / warn / disable decided per rule) so the baseline gate ends green.
 **Pitfalls this phase owns**:
  - Pick a baseline ruleset (recommended vs strict-type-checked) deliberately — strict surfaces a large upfront cleanup; decide blocking vs advisory before flipping the gate to blocking.
  - `typecheck`/tsc already gates type errors; ESLint should not duplicate type-checking rules unnecessarily.
 **Plans**: 3 plans — all complete (scope expanded during planning to add a Prettier `format:check` gate)
 - [x] 13-01-PLAN.md — Install ESLint/Prettier deps + flat config + package scripts + prove the gate fails (SC-1)
 - [x] 13-02-PLAN.md — Fix all first-run lint violations across both apps, green `pnpm lint` (D-13-05/06)
 - [x] 13-03-PLAN.md — Prettier reformat (isolated commit) + CI format:check step + green baseline (SC-2/SC-3)
 **UI hint**: no
 ### Phase 14: Desktop E2E Coverage
 **Goal**: The Phase 8 regression gate exercises the desktop layout and flows, not just mobile. A `desktop` Playwright project (`devices['Desktop Chrome']`, no touch, wide viewport) is added to `apps/pwa/playwright.config.ts`, and the existing mobile-authored specs are reviewed/adjusted (or appropriately skipped) so `pnpm test:e2e` passes on a no-touch desktop viewport as well as the `iphone`/`pixel` profiles.
 **Mode:** standard
 **Depends on**: Phase 7 (the harness it extends) and Phase 8 (CI runs `pnpm test:e2e` and picks up the new project automatically — no CI plumbing change needed beyond any desktop-profile runtime/wait). Independent of Phases 9–13.
 **Requirements**: TBD (promoted from backlog 999.15)
 **Success Criteria** (what must be TRUE):
  1. A `desktop` project exists in `playwright.config.ts` (Desktop Chrome, wide viewport, no `hasTouch`).
  2. The existing e2e specs pass (or are explicitly, justifiably skipped) on the desktop profile — touch-gesture / mobile-drawer / mobile-only-layout assumptions are handled.
  3. `pnpm test:e2e` in CI runs and gates on both mobile and desktop profiles (blocking-vs-advisory for desktop decided when planned).
 **Pitfalls this phase owns**:
  - The real work is the spec-compat pass, not CI plumbing — Phase 8 reused the Phase 7 harness unchanged, so the config addition is small but specs authored for touch/mobile need per-spec review.
  - Desktop WebKit is optional — the Apple member is already covered on mobile Safari via `iphone`; Desktop Chrome is likely sufficient for a shared/wall browser.
 **Plans**: 1 plan
 Plans:
 - [x] 14-01-PLAN.md — Add the `desktop` Playwright project, desktop-skip the two mobile-only layout assertions (+ D-04 parity), update spec/README docs, and prove `pnpm test:e2e` is green on iphone + pixel + desktop with a blocking CI gate.
 **UI hint**: no
 ### Phase 15: Doc-Only CI Skip
 **Goal**: Doc-only PRs to `main` merge without running the slow `harness` (Playwright e2e + dev-stack bring-up, ~5 min) and `api` (MariaDB integration) jobs, while `fast-checks` (Prettier `format:check` + markdown linting) still runs — and branch protection never deadlocks on a required check that never reports. Docs get a *fast but real* gate: format + lint, none of the slow code jobs.
 **Mode:** standard
 **Depends on**: Phase 8 (the `.gitea/workflows/ci.yml` it modifies) and Phase 13 (the `fast-checks` job + `format:check` step this extends). Independent of Phases 9–12.
 **Requirements**: TBD (promoted from backlog 999.17)
 **Success Criteria** (what must be TRUE):
  1. A doc-only PR to `main` (only `docs/` or `*.md` changed) skips the `api` and `harness` jobs but still runs `fast-checks`.
  2. A PR touching code runs `fast-checks`, `api`, and `harness` as today; a failure in any blocks the merge.
  3. Branch protection requires `CI / fast-checks` + an always-running `CI / gate` aggregate (passes when each heavy job is `success` OR `skipped`) — the direct `api`/`harness` requirements are dropped so a skipped heavy job never deadlocks the merge.
  4. `fast-checks` runs a markdown linter (markdownlint-cli2) over `**/*.md`; an introduced markdown-lint violation fails the gate, and the existing markdown baseline passes (violations fixed or rules configured) so the gate starts green.
 **Pitfalls this phase owns**:
  - **Required-check deadlock** — never path-filter a required context directly; a required job that never reports blocks the PR forever. The always-running `gate` job (`if: always()`, passes on `success`/`skipped`) is the only safe gating surface.
  - **Gitea skipped-status quirk** — Gitea may not emit a commit-status for a `skipped` job; rely on the always-running `gate`, not on marking `api`/`harness` skipped-but-required.
  - **Prettier vs markdownlint overlap** — Prettier already owns markdown *formatting*; scope markdownlint to *content* rules (heading increments, no broken/duplicate link refs, list/code-fence conventions) and disable its purely-stylistic rules that fight Prettier (e.g. line-length, list-indent), so the two don't conflict on the same `.md`.
  - **`.planning/*` is push-direct, never linted** — planning bookkeeping bypasses CI via the Unprotected file pattern, so markdownlint never sees it; scope the lint glob to `docs/` + repo-root/app `*.md` and exclude `.planning/**` (and any generated markdown) to avoid a baseline cleanup of churny bookkeeping files.
 **Plans**: 3 plans (3 waves)
 Plans:
 **Wave 1**
 - [x] 15-01-PLAN.md — markdownlint-cli2 + `.markdownlint-cli2.jsonc` + `md:lint` script + fast-checks step + fix 13 baseline violations (SC-4)
 **Wave 2** *(blocked on Wave 1 completion)*
 - [x] 15-02-PLAN.md — ci.yml: `changes` (dorny/paths-filter@v4) + conditional api/harness + always-running `gate` aggregate (SC-1/SC-2, SC-3 YAML)
 **Wave 3** *(blocked on Wave 2 completion)*
 - [x] 15-03-PLAN.md — operator branch-protection checkpoint (require `CI / fast-checks` + `CI / gate`, drop api/harness) + publish.yml comment update (SC-3)
 **UI hint**: no
 ### Phase 16: CI Dependency Audit, Security Checks & Image Hygiene
 **Goal**: The CI pipeline surfaces outdated and vulnerable dependencies, runs a baseline of additional security checks, and enforces a clean dev↔prod boundary in the images it publishes — so the two-person household app doesn't silently rot on stale/CVE-bearing packages, and no dev-only affordance, secret, or family-specific data ever ships in a production image. Extends the existing Gitea CI (Phase 8) workflow with dependency/security/image-hygiene gates rather than standing up a separate pipeline. **Absorbs backlog 999.17 (dev/prod image boundary).**
 **Mode:** standard
 **Depends on**: Phase 8 (Gitea CI — adds steps to the existing workflow + publish job; no admin-chain dependency). Independent of Phases 10–12.
 **Requirements**: SEC-01 (secret scanning), SEC-02 (static security lint), DEP-01 (vuln audit gate), DEP-02 (outdated advisory), IMG-01 (NODE_ENV+boot-guard), IMG-02 (.dockerignore), IMG-03 (publish image-hygiene assertions), CI-03 (security job + gate wiring)
 **Candidate scope (to be sharpened in `/gsd-discuss-phase 16`):**
  - **Outdated dependencies:** a CI step that reports dependencies behind their latest (e.g. `pnpm outdated -r`), surfaced on the PR. Decide gating vs advisory, and how to handle the pinned-version table in CLAUDE.md (the stack pins exact versions — "outdated" must not fight intentional pins).
  - **Vulnerability audit:** `pnpm audit` (or equivalent) against the lockfile, failing on a chosen severity threshold (e.g. high/critical). Decide the threshold and an allowlist/waiver mechanism for unfixable transitive advisories.
  - **Additional security checks (user is open to these — pick a sensible baseline, avoid over-build):** candidates — secret scanning on the diff (gitleaks/trufflehog), a CodeQL/`eslint-plugin-security` static pass, dependency-review on PRs, Dockerfile/image scan (e.g. trivy) of the published image.
  - **Dev/prod boundary definition & enforcement (from 999.17):** the `DEV_AUTH_BYPASS` concept (and any dev-only affordance) must be provably confined to local dev — never to production, never baked into published images. Today the guard is runtime-only (`NODE_ENV !== 'production' && DEV_AUTH_BYPASS === 'true'` in `apps/api/src/auth/devBypass.ts`); add (a) explicit documentation of what "dev image" vs "shipped image" means, and (b) build-time / boot-time enforcement (a `production` image refuses to boot — or the build aborts — if dev-bypass is enabled) as defense-in-depth.
  - **No data/secrets in published images (from 999.17):** audit the Dockerfile(s) + the Phase 8 publish job (`publish.yml`) to confirm `.env`, dev seed SQL, local DB dumps, encryption keys, OIDC secrets, the `DEV_USER` seed, and family-specific fixtures are `.dockerignore`d and never `COPY`'d. Add a CI assertion that fails the publish if a dev-bypass code path is active, a forbidden env/secret is present, or personal/seed data is staged into the image context. The dev-stack seed path (`DEV_USER` id 1 + sample calendar/list data) must be unreachable from the production image/compose.
  - **Noise control:** these gates are notorious for flaky/advisory-churn failures; decide blocking-on-merge vs warn-only per check, and where results surface (PR annotation vs job log), mirroring Phase 15's gate-aggregation approach.
 **Boundary:** Extends the existing Gitea CI workflow + publish job; does not remove dev-bypass (still needed for local verification and the Phase 7/8 harness) and does not add a new external service or a runtime dependency to the app. Automated dependency *upgrades* (e.g. Renovate/Dependabot bots) are a separate concern — decide in discuss whether they're in scope or deferred.
 **Plans**: 6 plans in 2 waves
 Plans:
 **Wave 1**
 - [x] 16-01-PLAN.md — Image-hygiene runtime: bake NODE_ENV=production + boot-time refuse-to-boot guard (IMG-01)
 - [x] 16-02-PLAN.md — pnpm audit gate + waiver allowlist + advisory-only tiered outdated report (DEP-01, DEP-02)
 - [x] 16-03-PLAN.md — Fold eslint-plugin-security into the lint gate as blocking errors + triage (SEC-02)
 - [x] 16-04-PLAN.md — gitleaks config + full-history baseline + .dockerignore (SEC-01, IMG-02)
 **Wave 2** *(blocked on Wave 1 completion)*
 - [x] 16-05-PLAN.md — Add the security job to ci.yml (gitleaks always; audit/outdated code-gated) + gate wiring (CI-03)
 - [x] 16-06-PLAN.md — publish.yml static image-hygiene assertion + boot-smoke before push (IMG-03)
 **UI hint**: no
 ### Phase 17: UI Optimization & Polish
 **Goal**: A visual-identity & polish pass for the PWA spanning three workstreams: **(A) phone-layout polish** so the phone (≤767px) layout has no fixed-chrome overlap and small-viewport spacing reads cleanly — starting with the long-standing BottomTabBar overlap that hides the New Event FAB and the colour legend, plus a small-viewport sweep; **(B) branding assets** — generate a real FamilySync logo into the existing `BrandSlot` seam (`apps/pwa/src/components/BrandSlot.tsx`) and a complete favicon/PWA-icon set replacing the placeholder stubs in `apps/pwa/public/`; **(C) theme-token groundwork** — restructure `apps/pwa/src/styles/tokens.css` into a themeable semantic-token layer (swappable by `data-theme`/`prefers-color-scheme`), light staying the only shipped theme, so a future dark theme is cheap.
 **Mode:** standard
 **Depends on**: Nothing structural (CSS/layout + assets only). Best sequenced after Phase 10 merges (the BottomTabBar gained an Admin tab and the new SetupBanner adds top pressure on phone), but otherwise independent of the admin chain.
 **Requirements**: No REQ-IDs — decisions D-01…D-10 (17-CONTEXT.md) stand in. Coverage: D-01/D-02 (A, phone overlap+guard) → 17-03; D-03/D-04 (B, assets) → 17-02; D-04/D-05 (B, wiring) → 17-04; D-06 (C, token groundwork) → 17-01; D-07/D-09 (D, logout+sheet centering) → 17-05; D-08/D-09/D-10 (D, admin toasts+reset-sheet+two-tab nav) → 17-06.
 **Scope boundary (set in `/gsd-discuss-phase 17`, 2026-06-17):** Workstream C ships token groundwork **only** — no dark palette, no theme toggle (→ backlog **999.20**). A broader "modern styling" visual refresh is **out of scope** and routed to backlog **999.21** (future milestone). Keep Phase 17 a focused polish + branding + groundwork pass, not a redesign.
 **Seed defect — phone-layout bottom-bar overlap (documented 2026-06-13; long-standing, NOT introduced by Phase 10 — the BottomTabBar dates to Phase 04):**
 At ≤767px (`window.matchMedia('(max-width: 767px)')` in `apps/pwa/src/App.tsx`) the layout switches to a 48px top AppNav + a `position: fixed` BottomTabBar (`height: calc(56px + env(safe-area-inset-bottom))`, z-index 200; `apps/pwa/src/components/BottomTabBar.tsx`) + a floating "New Event" FAB (`position: fixed; bottom: var(--space-6); right: var(--space-6)`; `apps/pwa/src/components/CalendarShell.tsx`). Two problems:
  1. **FAB sits inside the bar** — the FAB's `bottom` offset (~`--space-6`, ≈24px) is smaller than the bar's 56px height, so the round New Event button overlaps the bottom tab bar (lands on the Admin tab).
  2. **Content occluded** — the content area (`contentStyle` in `App.tsx`) reserves no `padding-bottom` for the fixed bar, so the bottom of the calendar and the colour-legend chips (e.g. the "Dev User" / member legend) slide under the bar and are partially hidden.
 **Fix sketch (CSS-only, no behaviour change):** on phone, lift the FAB to `bottom: calc(56px + env(safe-area-inset-bottom, 0px) + var(--space-6))` and add a matching `padding-bottom: calc(56px + env(safe-area-inset-bottom, 0px))` to the phone content/scroll area (or reduce the `100dvh` column by the bar height). Verify across the `iphone`/`pixel`/`desktop` Playwright profiles and a real narrow Chromium via playwright-cli.
 **Evidence:** reproduced 2026-06-13 with playwright-cli at 390×844 (FAB over the Admin tab; "Dev User" legend clipped) vs 1280×800 (desktop sidebar, no overlap). Full detail in todo `2026-06-13-pwa-phone-bottombar-overlap.md`.
 **Candidate scope (to sharpen in `/gsd-discuss-phase 17`):** the seed defect above, plus a sweep for other small-viewport spacing / tap-target / overlap issues (the Phase 7 `layout.spec.ts` tap-target/overflow assertions are a ready checklist) and any phone/desktop visual inconsistencies noticed in use. Keep it a focused polish pass, not a redesign.
 **Plans**: 6/6 plans complete
 Plans:
 **Wave 1**
 - [x] 17-01-PLAN.md — C: tokens.css themeable-layer groundwork + --bottom-chrome-h token (D-06) [Wave 1]
 - [x] 17-02-PLAN.md — B: generate logo + full icon set, operator approval checkpoint (D-03, D-04) [Wave 1]
 **Wave 2** *(blocked on Wave 1 completion)*
 - [x] 17-03-PLAN.md — A: phone FAB/BottomTabBar overlap fix + sweep + CI overlap assertion (D-01, D-02) [Wave 2, dep 01]
 - [x] 17-04-PLAN.md — B: wire logo into BrandSlot + index.html favicons + manifest maskable + accent (D-04, D-05) [Wave 2, dep 01,02]
 - [x] 17-05-PLAN.md — D: logout control + sheet desktop-centering (SettingsSheet/CredentialSheet) (D-07, D-09) [Wave 2, dep 01]
 - [x] 17-06-PLAN.md — D: admin success toasts + two-tab ARIA nav + reset-sheet centering + admin.spec.ts (D-08, D-09, D-10) [Wave 2, dep 01]
 **UI hint**: yes
 ## Progress
 | Phase | Milestone | Plans Complete | Status   | Completed  |
 | ----- | --------- | -------------- | -------- | ---------- |
 | 1. Foundation + Broker Spike     | v1.0 | 4/4   | Complete | 2026-06-04 |
 | 2. Calendar Display              | v1.0 | 5/5   | Complete | 2026-06-05 |
 | 3. Event Write-Back + PWA Install| v1.0 | 12/12 | Complete | 2026-06-07 |
 | 4. Shared Lists + Live Sync      | v1.0 | 7/7   | Complete | 2026-06-09 |
 | 5. Web Push Notifications        | v1.0 | 8/8   | Complete | 2026-06-10 |
 | 6. UX Polish                     | v1.0 | 6/6   | Complete | 2026-06-10 |
 | 7. Mobile Test Harness           | v1.1 | 4/4 | Complete    | 2026-06-11 |
 | 8. Gitea CI                      | v1.1 | 4/4 | Complete   | 2026-06-11 |
 | 9. Faster Write-Back             | v1.1 | 2/2 | Complete    | 2026-06-12 |
 | 10. Admin Role & Settings        | v1.1 | 4/4 | Complete    | 2026-06-13 |
 | 11. Per-Event Reminders          | v1.1 | 5/5 | Complete    | 2026-06-14 |
 | 12. Initial Setup Wizard         | v1.1 | 7/7 | Complete    | 2026-06-16 |
 | 13. Real Lint Gate (ESLint)      | v1.1 | 3/3   | Complete    | 2026-06-12 |
 | 14. Desktop E2E Coverage         | v1.1 | 1/1 | Complete    | 2026-06-12 |
 | 15. Doc-Only CI Skip + MD Lint   | v1.1 | 3/3 | Complete    | 2026-06-12 |
 | 16. CI Dep Audit, Sec & Img Hyg  | v1.1 | 6/6 | Complete    | 2026-06-13 |
 | 17. UI Optimization & Polish     | v1.1 | 6/6 | Complete    | 2026-06-18 |
 | 18. Auto Timezone Detection      | v1.1 | 4/4 | Complete    | 2026-06-14 |
 | 19. Local Auth (No-OIDC Mode)    | v1.1 | 5/5 | Complete    | 2026-06-17 |
 | 20. Admin Member Editor & Declutter | v1.1 | 3/3 | Complete    | 2026-06-18 |
 ## Backlog
 ### Phase 999.1: Treat Fastmail as one calendar provider; framework supports adding more providers (BACKLOG)
 **Goal:** [Captured for future planning] Abstract the calendar backend behind a provider interface so Fastmail/CalDAV is one implementation among potentially many. Shipping with a single provider is fine, but the broker, sync, and event-expansion layers should be structured so additional providers (e.g. other CalDAV hosts, Google Calendar, generic ICS feeds) can be added without rework. Captures the "provider" seam as an explicit architectural concern.
 **Requirements:** TBD
 **Plans:** 6/6 plans complete
 Plans:
 - [ ] TBD (promote with /gsd-review-backlog when ready)
 ### Phase 999.4: Per-event reminder configuration (VALARM authoring + scheduler honors it) (BACKLOG)
 **Goal:** [Captured for future planning] End-to-end per-event reminders — let the user choose *when* (or whether) to be reminded per event, and make the push scheduler honor that choice instead of a hardcoded lead.
 **Half A — author the VALARM (event form):** The event create/edit form has no UI to set a reminder ("remind me 10 min / 1 hour / 1 day before", or **no reminder**), so the written `.ics` carries no `VALARM` and no reminder can fire — in native clients or via web push. Add a reminder selector (including an explicit "none"), serialize chosen offsets as `VALARM` (TRIGGER) on write-back, and parse existing `VALARM`s on read so edits preserve them. Feeds the Phase 5 web-push requirement (push needs reminder data to notify about).
 **Half B — scheduler honors the provider's value (NEW, surfaced 2026-06-10):** Today `apps/api/src/broker/reminderScheduler.ts` runs a **hardcoded 15-minute** scan for shared timed events (`index.ts:139` "starting in ~15 min"; reminderScheduler header "15-min reminder scan") and never reads the event's actual alarm. So every reminder fires 15 min before regardless of what the event (or the calendar provider) specifies, and an event with **no** alarm still gets a 15-min push. Change the scheduler to read each event's `VALARM` `TRIGGER` (the value written in Half A / set in Fastmail or another native client) and fire at that lead — and fire **nothing** when the event has no alarm. The current fixed 15-min window/dedup logic (catch-up scan, per-uid exactly-once — see quick 260610-hbu) must be generalized to a variable per-event lead.
 **Boundary:** preserve the reminder scheduler's resilience guarantees (catch-up on a missed tick, per-uid exactly-once dedup). This makes the lead per-event/variable rather than constant; it is not a rewrite of the scan/dedup design.
 **Severity:** medium — feature gap surfaced during Phase 03 Gate 2 testing; Half B surfaced 2026-06-10. Tags: phase-03, phase-05, calendar, write-back, reminders, valarm, push, scheduler, phase-05-dependency.
 **Requirements:** TBD
 **Plans:** 0 plans
 > **Promoted into v1.1 Phase 11 (Per-Event Reminders) — CAL-13/CAL-14/NOTIF-04/05/06.** Backlog entry retained for history.
 Plans:
 - [ ] TBD (promote with /gsd-review-backlog when ready)
 ### Phase 999.10: Admin Settings / Administration section — manage app passwords + designate the shared calendar via UI (BACKLOG)
 **Goal:** [Captured for future planning] Add an in-app **Settings/Administration** section, gated to an administrator role, for configuration that today requires manual backend/DB steps:
 - **View/update per-member Fastmail app passwords** (stored encrypted via `APP_PASSWORD_ENCRYPTION_KEY`, existing crypto path) — rotate or re-enter a member's credential and re-trigger sync.
 - **Designate which synced calendar is the "shared" calendar** by toggling `calendars.is_shared` from the UI. Today this is a manual DB write: e.g. `UPDATE calendars SET is_shared=1 WHERE id=<row>` — done by hand on 2026-06-10 to mark the "FamilySync" calendar (id 10) shared after the poller synced it (D-16). The admin should pick the shared calendar from a list of synced collections instead of relying on a backend process. (The poller's upsert already leaves `is_shared` untouched, so a UI-set flag persists.)
 **Context:** Motivated by the manual D-16 resolution (2026-06-10). **Related:** 999.5 (per-member first-login app-password onboarding) — this is the ongoing admin-managed counterpart; and 999.11 (initial setup wizard) — bootstrap-time vs. ongoing config. Tags: admin, settings, calendar, app-passwords, D-16.
 > **Promoted into v1.1 Phase 10 (Admin Role & Settings) — ADMIN-01/ADMIN-02/ADMIN-03.** Backlog entry retained for history.
 **Requirements:** TBD
 **Plans:** 0 plans
 Plans:
 - [ ] TBD (promote with /gsd-review-backlog when ready)
 ### Phase 999.11: Initial setup wizard — first-run config of env vars, app passwords, DB connection (BACKLOG)
 **Goal:** [Captured for future planning] Add a first-run **setup wizard** that walks the administrator through defining all bootstrap configuration instead of hand-editing `.env` / `docker-compose.yml`:
 - **App environment variables:** OIDC client id/secret/issuer/redirect URI + external URL, session signing secret (`OIDC_AUTH_SECRET`), `APP_PASSWORD_ENCRYPTION_KEY`, and the **VAPID keypair** (subject + public + private).
 - **MariaDB connection:** host/port/user/password/db, with a connectivity test.
 - **First Fastmail app password** for the initial member, encrypted on save.
 Wizard should **validate inputs before completing** — e.g. VAPID private key decodes to 32 bytes AND pairs with the public key, OIDC discovery resolves, DB connects, app-password reaches CalDAV.
 **Context:** Motivated by setup friction observed 2026-06-10 — a VAPID private key truncated on paste into `.env` silently broke push (`setVapidDetails failed — 32 bytes`), and `DB_HOST` / dev overrides must currently be set by hand. A guided + validated wizard would have caught these. **Related:** 999.10 (ongoing admin Settings) and 999.5 (member onboarding). Tags: onboarding, setup, install, env, vapid, mariadb, oidc.
 > **Promoted into v1.1 Phase 12 (Initial Setup Wizard) — SETUP-01/02/03/04.** Backlog entry retained for history.
 **Requirements:** TBD
 **Plans:** 0 plans
 Plans:
 - [ ] TBD (promote with /gsd-review-backlog when ready)
 ### Phase 999.12: Assistant-driven mobile-browser UI testing (mobile viewport + authed PWA) (BACKLOG)
 **Goal:** [Captured for future planning] Give the assistant a way to validate UI/UX changes in a **mobile** browser experience, not just desktop Chromium. Today `playwright-cli` drives a desktop viewport, and the prod stack enforces OIDC (Authelia) so the authed PWA can't be reached headlessly — which is exactly why a string of mobile-only defects this milestone (silent Android notifications, the dead "How to enable" link, iOS/Android session-cookie persistence, install/standalone behaviour) could only be found by the operator on real devices, not by the assistant.
 **What this needs (any subset):**
 - **Mobile viewport + UA emulation** in the browser harness (e.g. Playwright device descriptors — iPhone/Pixel viewport, touch, mobile user-agent) so layout, tap targets, and responsive behaviour can be checked.
 - **An authenticated entry path for automated runs** so the assistant can reach the real PWA past Authelia — e.g. a reusable saved storage-state/cookie, a test-only bypass on a non-prod host, or driving the Authelia login once and reusing the session. (Note: this overlaps the existing `DEV_AUTH_BYPASS`, but that only works on the host-side dev stack, not the prod-mode PWA that has the real service worker. A mobile, authed, SW-enabled target is the gap.)
 - Optionally: a documented way to point the harness at the Pangolin HTTPS URL with a persisted session, and/or remote-debug a real device.
 **Boundary:** genuinely device-only behaviour (iOS-Safari standalone push, real APNs/FCM delivery, OS notification-channel importance) still needs a human — this item is about everything SHORT of that (responsive layout, tap flows, in-page notification UI states, auth redirects) which a mobile-emulated authed browser *could* cover but currently can't.
 **Context:** Surfaced 2026-06-10 during Phase 5 UAT — repeated mobile-only bugs were caught only by the operator because the assistant had no mobile, authenticated browser to test in. **Related:** [[feedback-playwright-verify]] (use playwright-cli over manual verification — this extends it to mobile/authed). Tags: testing, playwright, mobile, pwa, oidc, dx.
 > **Promoted into v1.1 Phase 7 (Mobile Test Harness) — TEST-01/TEST-02.** v1.1 scopes the `DEV_AUTH_BYPASS` dev-build path; the prod-SW authed-mobile target stays deferred. Backlog entry retained for history.
 **Requirements:** TBD
 **Plans:** 0 plans
 Plans:
 - [ ] TBD (promote with /gsd-review-backlog when ready)
 ### Phase 999.13: Reduce event write-back latency to the calendar provider (outbox drain) (BACKLOG)
 **Goal:** [Captured for future planning] Calendar create/edit/delete writes are enqueue-only (`calendarOutbox`, 202 optimistic-accept; D-12/D-05 — no Fastmail call in the route) and flushed to Fastmail by `runOutboxDrain` on a **15-second `setInterval`** (`apps/api/src/broker/outboxWorker.ts`). So a change can take up to ~15s to land in Fastmail (and longer to reflect back in the app, which depends on the separate 5-min poller). Reduce that perceived sync delay so edits feel near-immediate.
 **Options to weigh when picking this up:**
 - **Event-driven drain (preferred):** trigger an outbox drain immediately after a successful enqueue (in-process signal, or Redis pub/sub which is already available) so the write fires within ~1s instead of waiting for the next tick — keep the 15s `setInterval` as a fallback/retry sweep. Must preserve the existing per-row etag/412 handling and the rapid-successive-edit ordering (see outboxWorker comments ~L312 — each edit carries its enqueue-time etag).
 - **Shorter interval:** simplest, but more idle DB polling; a floor (e.g. 3–5s) trades latency for load.
 - **Faster read-back too:** the user also sees latency from the 5-min poller reflecting the change back. Consider invalidating/short-poll after a local write, or optimistic UI already covering it — confirm whether the perceived delay is the write (15s) or the read-back (5min).
 **Boundary:** the optimistic 202 + outbox durability design (create-before-delete, drain concurrency guard, fresh-etag-before-PUT) must be preserved — this is a latency tune, not a rewrite of the write path.
 **Context:** Surfaced 2026-06-10. Tags: calendar, write-back, outbox, latency, redis, performance.
 > **Promoted into v1.1 Phase 9 (Faster Write-Back) — CAL-15.** In-process EventEmitter chosen (not Redis); the drain is single-process by design. Backlog entry retained for history.
 **Requirements:** TBD
 **Plans:** 0 plans
 Plans:
 - [ ] TBD (promote with /gsd-review-backlog when ready)
 ### Phase 999.14: Gitea CI — full regression on PR to main + build/publish Docker image (BACKLOG)
 **Goal:** [Captured for future planning] The repo is committed against a self-hosted Gitea instance with a registered Actions runner, but there is no CI yet (no `.gitea/workflows/` or `.github/workflows/`). Two things should run automatically: (1) **full regression** on every PR targeting `main` — gating the merge; (2) **build the app's Docker image and publish it** to the Gitea container registry.
 **Options / decisions to make when picking this up:**
 - **Test scope:** "full regression" = lint + typecheck + unit + the API integration tests. Integration tests need a real MariaDB (see [[api-integration-test-db]]) — the workflow must spin up a MariaDB service container, bind it, and set `DB_HOST=127.0.0.1` + `.env` creds. The PWA build/test also runs.
 - **Monorepo:** pnpm workspace (`apps/api`, `apps/pwa`, shared). Cache the pnpm store.
 - **Docker images:** only `apps/api/Dockerfile` exists today — there is no PWA Dockerfile yet. Decide one image (API) vs. also building/serving the PWA. Tag scheme + when to publish (only on merge to `main`? on tags? per-PR?).
 - **Registry auth:** push to the Gitea registry using the runner's Gitea-provided token or a dedicated package-write token.
 - Gitea Actions are GitHub-Actions-compatible syntax but run on the self-hosted runner — confirm runner labels and available images, and that Actions is enabled, before authoring.
 **Likely shape:** a `.gitea/workflows/ci.yml` — `on: pull_request` (to `main`) → install (pnpm), lint, typecheck, unit, API integration vs. a `mariadb` service container, PWA build; `on: push` to `main`/tag → `docker build apps/api/Dockerfile`, login, push tagged image.
 **Context:** Promoted from STATE.md pending todo (`.planning/todos/pending/2026-06-10-gitea-ci-regression-and-docker-publish.md`), surfaced 2026-06-10. Tags: tooling, ci, gitea, docker, mariadb, monorepo.
 > **Promoted into v1.1 Phase 8 (Gitea CI) — CI-01/CI-02.** v1.1 also extends CI-01 to run the Phase 7 mobile harness as a UI-regression step (CI brings up the dev stack in the runner). Backlog entry retained for history.
 **Requirements:** TBD
 **Plans:** 0 plans
 Plans:
 - [ ] TBD (promote with /gsd-review-backlog when ready)
 ### Phase 999.15: Desktop e2e coverage — add a Desktop Playwright profile + desktop-safe specs (BACKLOG)
 **Goal:** [Captured for future planning] The Playwright harness (`apps/pwa/playwright.config.ts`) defines only **mobile** device profiles — `iphone` (iPhone 14 / WebKit) and `pixel` (Pixel 7 / Chromium), both with touch and a mobile viewport. The Phase 8 CI regression gate runs `pnpm test:e2e`, so it currently validates the **mobile experience only**. Add desktop coverage so the regression gate exercises the desktop layout/flows as well.
 **Options / decisions to make when picking this up:**
 - **Add a Desktop profile:** a new `desktop` project in `playwright.config.ts` (e.g. `devices['Desktop Chrome']`, no `hasTouch`, wide viewport). Optionally a Desktop WebKit/Safari profile too — but the family's Apple member is already covered on mobile Safari via `iphone`; Desktop Chrome is likely sufficient for a shared/wall browser.
 - **Spec-compat pass (the real work):** the existing e2e specs were authored for mobile — they may assume touch gestures, a mobile nav/drawer, or mobile-only layout. Each spec needs review/adjustment so it passes (or is appropriately skipped) on a no-touch, wide-viewport desktop. This is harness/spec work, not CI plumbing.
 - **Gating choice:** decide whether desktop runs block the merge immediately, or run advisory (non-blocking) until the specs are confirmed desktop-safe.
 **Boundary:** Phase 8 deliberately reused the Phase 7 harness **unchanged** (CI owns only stack bring-up + readiness waits, not spec content), which is why this was deferred. Once a Desktop project is added to the config, Phase 8 CI picks it up automatically via `pnpm test:e2e` — no CI changes needed beyond whatever runtime/wait the desktop profile requires.
 **Context:** Deferred from Phase 8 (Gitea CI) planning, 2026-06-11 — user wants both mobile and desktop validated, but desktop needs a config addition + spec review that is out of Phase 8's CI-plumbing scope. Tags: testing, playwright, e2e, desktop, harness, ci.
 > **Promoted into v1.1 Phase 14 (Desktop E2E Coverage) — 2026-06-11.** Backlog entry retained for history.
 **Requirements:** TBD
 **Plans:** 0 plans
 Plans:
 - [ ] TBD (promote with /gsd-review-backlog when ready)
 ### Phase 999.16: Wire a real linter (ESLint) so the CI lint gate actually fails on violations (BACKLOG)
 **Goal:** [Captured for future planning] The Phase 8 CI `fast-checks` job runs `pnpm lint`, but **no linter exists** in the repo — the root `lint` script is `pnpm -r --if-present lint`, which finds no package-level lint script and exits 0. The lint gate is a hollow placeholder that can never fail. Wire up a real linter so it runs and gates merges on lint violations. (`typecheck`/tsc already gates type errors meanwhile.)
 **Options / decisions to make when picking this up:**
 - **Tooling:** ESLint flat config (`eslint.config.js`) with `typescript-eslint`; add React + react-hooks plugins for `apps/pwa`. Add `eslint` (+ plugins) as devDeps and a `lint` script to `apps/api` and `apps/pwa` — `pnpm -r --if-present lint` then picks them up automatically, no CI change needed.
 - **Rule strictness:** pick a baseline (recommended vs strict-type-checked). Stricter = more upfront violations to fix.
 - **Violation cleanup (the real work):** the first run surfaces existing violations across both apps. Decide per-rule: fix, downgrade to warn, or disable. The gate must end green.
 - **Gating choice:** blocking on merge immediately, or advisory (warn-only) until the codebase is clean.
 **Boundary:** Phase 8 deliberately scoped lint wiring out (CI-plumbing-only); it shipped the gate slot wired to auto-activate once a package `lint` script lands. This item is that follow-up.
 **Context:** Raised during Phase 8 execution, 2026-06-11 — user noted the `--if-present` lint step "didn't fix the linter, just made it so it didn't have to exist to proceed" and wants a lint gate that actually fails. Tags: ci, lint, eslint, typescript-eslint, quality, gitea.
 > **Promoted into v1.1 Phase 13 (Real Lint Gate / ESLint) — 2026-06-11.** Backlog entry retained for history.
 **Requirements:** TBD
 **Plans:** 0 plans
 Plans:
 - [ ] TBD (promote with /gsd-review-backlog when ready)
 ### Phase 999.18: Update dependencies as found during CI (BACKLOG)
 **Goal:** [Captured for future planning] When the CI dependency-audit gate (Phase 16) surfaces outdated or vulnerable packages, bump them rather than letting the report accumulate. Establish a lightweight, recurring "act on the CI dependency report" loop so the two-person household app doesn't drift onto stale/CVE-bearing deps. Scope is the upkeep workflow (review → bump → verify gate green), not a one-time audit.
 **Context:** Captured 2026-06-13 during Phase 10 work. Companion to the audit *reporting* shipped in Phase 16 (CI Dependency Audit) — that phase makes outdated/vulnerable deps *visible*; this item is the standing follow-through to *resolve* what it finds. Tags: ci, dependencies, maintenance, security, upkeep.
 **Requirements:** TBD
 **Plans:** 0 plans
 Plans:
 - [ ] TBD (promote with /gsd-review-backlog when ready)
 ### Phase 999.19: Dev user exercises full app functionality without syncing to a real calendar (BACKLOG)
 **Goal:** [Captured for future planning] Let the `DEV_AUTH_BYPASS` dev user (currently hardcoded `DEV_USER` id 1 in `apps/api/src/auth/devBypass.ts`) exercise the full app — create/edit/delete events, set per-event reminders, manage lists — against a local/in-app calendar store, WITHOUT requiring a connected Fastmail/CalDAV provider and WITHOUT writing anything to a real calendar. Today the dev user has no `member_credentials` row and no `calendars`, so `writable-calendars` is empty and `POST /api/events/create` returns `422 "No writable calendar found for user"` — making hands-on UAT of event/reminder features impossible in dev. Options to explore: seed the dev user a fake local calendar + short-circuit the outbox/CalDAV write path under dev-bypass (no Fastmail round-trip), or a dev-only in-memory calendar provider. Must stay strictly dev-only (same hard `NODE_ENV !== 'production'` guard) and never ship in production images.
 **Context:** Captured 2026-06-14 during Phase 11 (Per-Event Reminders) UAT. The reminder picker and backend were verified via automated tests + a route-mocked playwright smoke, but the operator could not manually create an event to see reminders end-to-end because no provider is connected in the dev DB (`needsProviderSetup: true`). This is a recurring dev-testability friction (see MEMORY: "Dev user 1 has no calendars"). Tags: dev-tooling, dev-bypass, testability, calendars, outbox, uat.
 **Requirements:** TBD
 **Plans:** 0 plans
 Plans:
 - [ ] TBD (promote with /gsd-review-backlog when ready)
 ### Phase 999.20: PWA dark mode / theming — ship a full dark theme + light/dark/system switch (BACKLOG)
 **Goal:** [Captured for future planning] Ship a complete dark theme for the PWA plus a light/dark/system theme switch. **Phase 17 lays the token-architecture groundwork** — it restructures `apps/pwa/src/styles/tokens.css` from a single light `:root` into a themeable semantic-token layer that can be swapped via `data-theme` / `prefers-color-scheme`, with light staying the default and only-shipped theme. This backlog item is the follow-through that consumes that seam: author the actual dark palette values (including the Schedule-X `--sx-color-*` calendar overrides at the bottom of tokens.css), wire `prefers-color-scheme`, add a persisted in-app toggle in the /admin or Settings surface (light / dark / system), and verify both themes render cleanly across every route (calendar, lists, admin, settings sheet, login) via `playwright-cli` + the Phase 7 `layout.spec` profiles.
 **Context:** Deferred out of Phase 17 (2026-06-17) during `/gsd-discuss-phase 17` to keep that phase scoped to phone-layout polish + branding assets. Phase 17's token restructure is the explicit enabling groundwork, so this should be cheap to pick up afterward. Related: Phase 17 (UI Optimization & Polish — the groundwork), 999.21 (modern styling refresh). Tags: pwa, theming, dark-mode, tokens, accessibility, settings, prefers-color-scheme.
 **Requirements:** TBD
 **Plans:** 0 plans
 Plans:
 - [ ] TBD (promote with /gsd-review-backlog when ready)
 ### Phase 999.21: PWA modern visual styling refresh — contemporary look across the app (BACKLOG)
 **Goal:** [Captured for future planning] A broader "more modern, visually appealing" styling pass across the PWA — beyond the bounded in-system polish of Phase 17. Candidate scope: a contemporary refresh of high-visibility surfaces (login, calendar shell, event form, lists, admin), revisiting elevation/shadows, radii, spacing rhythm, typography scale, and control states, potentially reworking specific component layouts. Explicitly **flagged for a future milestone**, not v1.1 — it is a visual-overhaul track with real redesign risk and should be scoped/sequenced on its own rather than bolted onto a polish phase. Best sequenced after the Phase 17 token groundwork and 999.20 (dark mode) so the refresh is theme-aware from the start.
 **Context:** Deferred out of Phase 17 (2026-06-17) during `/gsd-discuss-phase 17`. The user scoped Phase 17 to layout polish + branding (logo/favicon/icon assets) + theme-token groundwork, and routed the open-ended styling refresh here for a future milestone to avoid an unbounded redesign inside a polish phase. Related: Phase 17 (the polish baseline), 999.20 (dark mode / theming). Tags: pwa, ui, styling, redesign, design-system, future-milestone.
 **Requirements:** TBD
 **Plans:** 0 plans
 Plans:
 - [ ] TBD (promote with /gsd-review-backlog when ready)
 ### Phase 18: Auto timezone detection and ability to change timezone
 **Goal:** Make the household timezone an explicit, stored, user-changeable setting — auto-detected from the browser at first run, changeable from the role-gated /admin Settings — and route the server-side all-day "9 AM local" reminder computation through it (replacing the implicit `process.env.TZ` fallback), without touching the already-correct browser-local display/timed-write path.
 **Requirements**: TBD (decision contract D-01..D-07 from 18-CONTEXT.md)
 **Depends on:** Phase 10 (admin role + `/admin` Settings + `app_config`); Phase 11 (all-day reminder computation this rewires). Independent of Phase 17. Phase 12 (setup wizard) not required — seeding is self-contained.
 **Plans:** 4/4 plans complete
 Plans:
 **Wave 1**
 - [x] 18-01-PLAN.md — TDD: getHouseholdTimezone(db) accessor + isValidIanaTimezone (D-05/D-06)
 **Wave 2** *(blocked on Wave 1 completion)*
 - [x] 18-02-PLAN.md — TDD: admin GET/PUT/seed timezone endpoints on adminRouter, requireAdmin + IANA validation + no-overwrite seed (D-01/D-02/D-03/D-04)
 - [x] 18-03-PLAN.md — TDD: route all-day reminder TZ at reminderScheduler:247 + outboxWorker:501,607 through the accessor (D-05/D-06/D-07)
 **Wave 3** *(blocked on Wave 2 completion)*
 - [x] 18-04-PLAN.md — PWA Timezone section in /admin Settings (searchable IANA picker + detected-zone seed) + client fns (D-02/D-04)
 **UI hint**: yes
 ### Phase 19: Local Auth (No-OIDC Mode)
 **Goal:** Let an operator run FamilySync entirely on **local DB users with no OIDC** — username/password accounts and a local login flow that coexists with the Authelia OIDC path — and **optionally wire OIDC in later** by claiming/linking an existing local user to an OIDC identity. Removes the hard dependency on a deployed Authelia for small/solo self-hosters.
 **Mode:** standard
 **Depends on:** Phase 12 (Initial Setup Wizard) — builds directly on the pre-OIDC **local-user foundation** introduced there: nullable `users.oidc_iss`/`oidc_sub` + the claimed/pending marker, and the first-login-claims merge. Phase 19 generalizes that single bootstrap local user into a full local-account model + login.
 **Requirements**: AUTH-LOCAL-01..AUTH-LOCAL-20 (derived during planning 2026-06-17) — local_credentials schema (01), scrypt hash/verify (02), login route (03), localAuthMiddleware (04), auth-mode endpoint (05), logout (06), admin create-member (07), admin reset (08), self-change (09), OIDC-link (10), break-glass CLI (11), LoginPage (12), admin UI (13), settings UI (14), routing gate (15), dev-bypass/harness rework (16), hasLocalCredential (17), de-Authelia copy (18), rate-limit/lockout (19), auth unit tests (20). Plus `LOCAL_SESSION_SECRET` env + boot assertion (D-05).
 **Plans:** 5/5 plans complete
 **Provenance:** Deferred from the Phase 12 discussion (2026-06-15) — see `.planning/phases/12-initial-setup-wizard/12-CONTEXT.md` §Deferred Ideas. The operator runs FamilySync this way themselves and wants no-OIDC operation as a first-class mode.
 **Open questions for discuss/spec:**
 - Password hashing/storage choice (e.g. argon2id/bcrypt) and how it sits alongside the env-only secret kernel from Phase 12.
 - How local login coexists with `oidcAuthMiddleware` ordering in `apps/api/src/index.ts` (route-level auth strategy selection vs. a mode flag in `app_config`).
 - The OIDC-link flow: claiming an existing local user into an `oidc_iss+oidc_sub` identity without violating the D-10 "identity is OIDC, never email" rule.
 - Whether "local mode vs OIDC mode" is a deploy-time switch or both can be live simultaneously.
 Plans:
 **Wave 1**
 - [x] 19-01-PLAN.md — Foundation (TDD): local_credentials schema + 0003 migration, scrypt hash/verify, local-session JWT helpers, LOCAL_SESSION_SECRET boot guard + generate-secrets, .dockerignore scripts exclusion (AUTH-LOCAL-01/02)
 **Wave 2** *(blocked on Wave 1)*
 - [x] 19-02-PLAN.md — Backend account mgmt (TDD): admin create/reset member, self-change password, hasLocalCredential, linkOidcToUser helper + /api/me/link-oidc (AUTH-LOCAL-07/08/09/10/17)
 **Wave 3** *(blocked on Wave 2)*
 - [x] 19-03-PLAN.md — Middleware + routes + wiring (TDD): localAuthMiddleware, /api/auth/mode, login (rate-limit/lockout) + logout, index.ts mount + OIDC-guard skip + /callback link branch, de-Authelia comments (AUTH-LOCAL-03/04/05/06/18/19/20)
 **Wave 4** *(blocked on Wave 3; 04 + 05 parallel)*
 - [x] 19-04-PLAN.md — PWA: LoginPage + BrandSlot + App.tsx gate + client.ts + AdminPage + SettingsSheet (AUTH-LOCAL-12/13/14/15)
 - [x] 19-05-PLAN.md — Dev-bypass Option C + break-glass CLI + harness/CI rework + login.spec.ts (AUTH-LOCAL-11/16)
 ### Phase 20: Admin Member Editor & Form Declutter
 **Goal:** Replace the per-member-row action buttons (Rotate/Add credential + Reset password) in the admin Members panel with a single edit affordance — clicking a member's name or an edit button opens a member-detail editor where an admin modifies all of that member's details in one place: display name, local-login password, and the Fastmail/CalDAV app password (calendar credential) — using clear, non-jargon labels that retire the confusing "Rotate" term. Also collapse the "Add member" section so its input fields are hidden behind a single "Add member" trigger by default, decluttering the panel. Client-side AdminPage + CredentialSheet rework over the existing `/api/admin` endpoints; no new auth/authorization boundary (seeded by the gripe that "Rotate" for the app password is not intuitive).
 **Requirements**: TBD (refine in /gsd-discuss-phase 20 — open scope: which fields count as "all" (color swatch? admin toggle? OIDC link?), whether to keep any standalone reset-password flow, and the exact edit affordance — clickable name vs. row edit button)
 **Depends on:** Phase 19
 **Plans:** 3/3 plans complete
 Plans:
 **Wave 1**
 - [x] 20-01-PLAN.md — Server: PATCH /api/admin/members/:id (displayName + is_admin) with last-admin demotion guard (TDD) + isAdmin in GET /members
 - [x] 20-02-PLAN.md — PWA API client: AdminMember.isAdmin field + updateMemberProfile fetcher (last-admin sentinel)
 **Wave 2** *(blocked on Wave 1 completion)*
 - [x] 20-03-PLAN.md — PWA: unified MemberEditorSheet (edit/create, per-section saves) + decluttered tappable Members panel; retire Rotate/Reset-password buttons
@@ -0,0 +1,62 @@
 ---
 status: complete
 phase: 20-admin-member-editor-form-declutter
 source: [20-01-SUMMARY.md, 20-02-SUMMARY.md, 20-03-SUMMARY.md]
 started: 2026-06-19T00:28:25Z
 updated: 2026-06-19T00:34:38Z
 verified_by: playwright-cli (Chromium, http://localhost:5173/admin, dev stack)
 ---
 ## Current Test
 [testing complete]
 ## Tests
 ### 1. Open the Member Editor from a member row
 expected: Members panel shows tappable rows (name + chevron, no per-row action buttons). Tapping a row opens one Member Editor sheet with all of that member's details.
 result: pass
 evidence: Members panel rendered one tappable button "Edit Dev User" (name + "Credential set" + Admin badge + chevron) and a single "Add member" trigger. Tapping it opened the "Edit member" dialog containing Profile (display name + Admin switch), "Set new password", and "App password" sections in one sheet.
 ### 2. Edit display name + admin flag (Profile section)
 expected: Changing the display name and/or admin toggle and saving the Profile section persists; the row reflects the new name. No "saved" toast on a no-op save.
 result: pass
 evidence: Changed display name to "Dev User QA" → Save → GET /api/admin/members returned displayName "Dev User QA"; sheet stayed open (per-section save). Reverted to "Dev User" and re-verified.
 ### 3. Last-admin demotion is blocked
 expected: Demoting the only admin shows an inline error and reverts the toggle; member stays admin.
 result: pass
 evidence: Toggled Admin off (aria-checked=false) → Save → 409 from PATCH /api/admin/members/1; inline error "Cannot remove admin — at least one admin must remain." rendered; switch reverted to aria-checked=true; member still isAdmin=true in DB. (The single console error is the expected 409 — not a bug.)
 ### 4. Set a new local-login password (write-only)
 expected: Write-only password field (blank, never prefilled); does not echo any existing password.
 result: pass
 evidence: "Set new password" section fields start blank; New/Confirm/App-password inputs are type="password"; New-password autocomplete="new-password"; "Leave blank to keep the current password." helper shown. Save button disabled until filled.
 ### 5. Set / update the Fastmail app password (clear labels)
 expected: Clear non-jargon labels (no "Rotate"); valid password CalDAV-validated before store; email field blank on edit.
 result: pass
 evidence: Section labeled "App password" with helper "Fastmail app password scoped to Calendars & Contacts (CalDAV)." plus a "Get an app password" link. App-password and Fastmail-email fields both start blank (value=""). No "Rotate" term present.
 ### 6. Add a member via the collapsed trigger
 expected: "Add member" is a single collapsed trigger; tapping opens the editor in create mode.
 result: pass
 evidence: No inline always-open add form. Tapping the single "Add member" button opened an "Add member" dialog in create mode (Display name, Username, Initial password, Confirm password + disabled "Add member" submit).
 ### 7. Old jargon and buttons are gone
 expected: "Rotate", "Add credential", and the standalone "Reset password" button no longer appear.
 result: pass
 evidence: DOM innerText scan on /admin returned {rotate:false, addCredential:false, resetPassword:false}.
 ## Summary
 total: 7
 passed: 7
 issues: 0
 pending: 0
 skipped: 0
 blocked: 0
 ## Gaps
 [none — all tests passed]
@@ -0,0 +1,83 @@
 ---
 quick_id: 260618-smr
 slug: remove-unused-redis-service-and-referenc
 description: Remove unused Redis service and references
 type: quick
 created: 2026-06-19
 files_modified:
  - docker-compose.yml
  - docker-compose.dev.yml
  - CLAUDE.md
  - README.md
  - docs/ARCHITECTURE.md
  - docs/CONFIGURATION.md
  - docs/deployment.md
  - docs/DEVELOPMENT.md
  - docs/GETTING-STARTED.md
  - docs/TESTING.md
  - apps/pwa/e2e/README.md
  - apps/pwa/playwright.config.ts
 ---
 # Quick Task 260618-smr: Remove unused Redis service and references
 ## Why
 Redis is confirmed **unused at runtime**: no `ioredis`/redis client import, no `REDIS_*`
 env vars read in code, and `ioredis` is not a dependency in any `package.json`. It exists
 only as a compose service + documentation references that imply it is part of the stack or
 "reserved for future pub/sub". Decision: drop the container and all references (keep the
 in-memory-vs-Redis design-rationale comments — see Constraint below).
 ## Tasks
 ### Task 1 — Remove the redis service from compose
 - `docker-compose.yml`: delete the `redis:` service block (`image: redis:7-alpine` + its
  comment) so only `api`, `mariadb`, and the `volumes:` block remain. The api `depends_on`
  lists only `mariadb` — leave it untouched.
 - `docker-compose.dev.yml`: delete the `redis:` ports override block (`ports: - '6379:6379'`).
 - verify: `grep -ri redis docker-compose.yml docker-compose.dev.yml` returns nothing.
 - done: neither compose file references redis; `docker compose config` still parses.
 ### Task 2 — Clean documentation references
 Remove/adjust every Redis mention so no doc implies Redis is part of the stack:
 - `CLAUDE.md`: delete the Constraints line "Redis available (optional, …)"; delete the
  `ioredis` row from the Supporting Libraries table; delete the "Redis is present in the
  stack but not yet used…" sentence from the architecture paragraph (keep the rest of the
  sentence about SSE/EventEmitter); change the compose-tree comment `(api + mariadb + redis)`
  → `(api + mariadb)`.
 - `README.md`: drop "Redis" from the prerequisites line; drop `redis` from the
  `docker compose … up mariadb redis` command; drop ", Redis 7" from the compose-file
  description; change "expose DB/Redis ports" → "expose DB ports"; change the Live-sync row
  "Server-Sent Events + Redis 7 pub/sub" → "Server-Sent Events (in-process EventEmitter)".
 - `docs/ARCHITECTURE.md`: delete the `Redis` table row.
 - `docs/CONFIGURATION.md`: drop "and Redis on `localhost:6379`".
 - `docs/deployment.md`: delete the `redis` services-table row.
 - `docs/DEVELOPMENT.md`: remove the four Redis mentions (prereq bullet, "### 2. Start the dev
  database and Redis" heading → "Start the dev database", the two `up mariadb redis` commands
  → `up mariadb`, the "Exposes Redis on `localhost:6379`" bullet, and the
  "(API in Docker + MariaDB + Redis…)" comment → "(API in Docker + MariaDB…)").
 - `docs/GETTING-STARTED.md`: prereq row "Used to run MariaDB and Redis locally" → "MariaDB";
  `up -d mariadb redis` → `up -d mariadb`; "and Redis (`localhost:6379`)" removed from prose.
 - `docs/TESTING.md`: "The API, MariaDB, and Redis must already be running" → "The API and
  MariaDB must already be running".
 - `apps/pwa/e2e/README.md`: delete the "- Redis on `:6379`" bullet.
 - `apps/pwa/playwright.config.ts`: update the two comments listing
  "API+MariaDB+Redis are compose-managed" → "API+MariaDB are compose-managed".
 - verify: `grep -rniE redis CLAUDE.md README.md docs apps/pwa/e2e apps/pwa/playwright.config.ts`
  returns nothing (case-insensitive, excluding the word "credential").
 - done: no doc/config implies Redis is in the stack.
 ## Constraint — KEEP these (do NOT touch)
 The in-memory-vs-Redis **design-rationale** comments document why in-memory is used instead
 of Redis (decisions D-12/D-18) and must remain:
 - `apps/api/src/lib/listEmitter.ts`
 - `apps/api/src/broker/reminderScheduler.ts`
 - `apps/api/src/auth/linkNonceStore.ts`
 - `apps/api/src/routes/localAuth.ts`
 ## must_haves
 - truth: "No redis service exists in either compose file"
 - truth: "No documentation or e2e config references Redis as part of the stack"
 - truth: "The D-12/D-18 in-memory-vs-Redis rationale comments in the 4 source files are intact"
 - artifacts: [docker-compose.yml, docker-compose.dev.yml, CLAUDE.md, README.md, docs/*, apps/pwa/e2e/README.md, apps/pwa/playwright.config.ts]
@@ -0,0 +1,65 @@
 ---
 quick_id: 260618-smr
 slug: remove-unused-redis-service-and-referenc
 description: Remove unused Redis service and references
 status: complete
 completed: 2026-06-18
 duration: ~10m
 tasks_completed: 2
 tasks_total: 2
 files_modified:
  - docker-compose.yml
  - docker-compose.dev.yml
  - CLAUDE.md
  - README.md
  - docs/ARCHITECTURE.md
  - docs/CONFIGURATION.md
  - docs/deployment.md
  - docs/DEVELOPMENT.md
  - docs/GETTING-STARTED.md
  - docs/TESTING.md
  - apps/pwa/e2e/README.md
  - apps/pwa/playwright.config.ts
 decisions:
  - Redis dropped entirely from compose and all docs; in-memory EventEmitter is the live-sync mechanism (D-12/D-18 rationale comments in source preserved)
 ---
 # Quick Task 260618-smr: Remove unused Redis service and references — Summary
 ## One-liner
 Dropped the unused `redis:7-alpine` compose service and scrubbed every documentation/config reference implying Redis is part of the stack, while preserving the D-12/D-18 in-memory-vs-Redis design-rationale comments in source.
 ## Tasks Completed
 | Task | Description                          | Commit  | Files                                                      |
 | ---- | ------------------------------------ | ------- | ---------------------------------------------------------- |
 | 1    | Remove redis service from compose    | 269e474 | docker-compose.yml, docker-compose.dev.yml                 |
 | 2    | Clean documentation references       | 8255be6 | CLAUDE.md, README.md, docs/*, apps/pwa/e2e/README.md, apps/pwa/playwright.config.ts |
 ## Verification Results
 - `grep -ri redis docker-compose.yml docker-compose.dev.yml` → no output (clean)
 - `docker compose -f docker-compose.yml -f docker-compose.dev.yml config` → PARSE OK
 - `grep -rniE redis CLAUDE.md README.md docs apps/pwa/e2e apps/pwa/playwright.config.ts` → no output (clean)
 - `pnpm format:check` → all matched files use Prettier code style
 - Protected source files (listEmitter.ts, reminderScheduler.ts, linkNonceStore.ts, localAuth.ts) → untouched (git diff confirms no changes)
 ## Docker Compose Config
 `docker compose -f docker-compose.yml -f docker-compose.dev.yml config` parsed successfully with no Redis service — confirmed available on this host.
 ## Deviations from Plan
 None — plan executed exactly as written.
 ## Known Stubs
 None.
 ## Self-Check: PASSED
 - Commits 269e474 and 8255be6 exist in git log
 - All 12 modified files confirmed updated
 - Grep verifies zero Redis references in target files
 - Protected source files confirmed untouched
@@ -0,0 +1,99 @@
 ---
 quick_id: 260618-tg2
 slug: persistent-ci-dependency-caches-pnpm-sto
 description: Persistent CI dependency caches (pnpm store + Playwright browsers)
 type: quick
 created: 2026-06-19
 files_modified:
  - .gitea/workflows/ci.yml
  - docs/DEVELOPMENT.md
  - apps/api/Dockerfile
  - .gitea/workflows/publish.yml
 ---
 # Quick Task 260618-tg2: Persistent CI dependency caches (pnpm store + Playwright)
 ## Why
 The Gitea runner re-downloads all deps every run: 4 jobs each run `pnpm install --frozen-lockfile`
 cold (lines 54/110/197/478), and the harness job re-downloads Playwright browser binaries every
 run (line 275). The runner is long-lived Docker-on-Unraid, so persisting these via host bind-mounts
 (`/pnpm-store`, `/ms-playwright`, wired in the act_runner `config.yaml` `container.options` — a
 separate manual host change) eliminates the repeat downloads. This avoids `actions/cache@v4`, which
 the Phase-8 runner probe found times out on this runner (D-PROBE-04).
 **Scope this task: the two CI caches only.** Verdaccio (registry mirror) and the Dockerfile
 BuildKit cache mount are explicitly OUT of scope for now.
 ## Tasks
 ### Task 1 — pnpm store: point all CI installs at the persistent store
 In `.gitea/workflows/ci.yml`, change each of the four install steps:
 ```
 run: pnpm install --frozen-lockfile
 ```
 →
 ```
 run: pnpm install --frozen-lockfile --store-dir /pnpm-store --prefer-offline
 ```
 Lines 54 (fast-checks), 110 (api), 197 (harness), 478 (security). `--store-dir` and
 `--prefer-offline` are valid pnpm 11.5.1 install flags (verify with `pnpm install --help`).
 Do NOT add `store-dir` to a repo `.npmrc` — local dev has no `/pnpm-store`.
 Update the stale comment near line 51 (the "no cache backend / ~30s acceptable" note) to reflect
 that installs now use the persistent host-mounted store.
 - verify: `grep -c -- '--store-dir /pnpm-store --prefer-offline' .gitea/workflows/ci.yml` → 4
 - done: all four installs use the persistent store; YAML still valid.
 ### Task 2 — Playwright: persist browser binaries on the harness job
 Add a job-level `env:` to the `harness:` job so every step (browser install + test run) resolves
 the same persistent path:
 ```yaml
  harness:
    runs-on: ubuntu-latest
    env:
      PLAYWRIGHT_BROWSERS_PATH: /ms-playwright
 ```
 (If the harness job already has a job-level `env:` map, add the key to it rather than duplicating.)
 Leave `npx playwright install --with-deps webkit chromium` (line 275) as-is — the binary download
 is now cached by version; the `--with-deps` apt step can't persist (add a one-line comment noting
 "baking a runner image with browsers preinstalled would also drop the --with-deps apt step" as a
 future optimization).
 - verify: `PLAYWRIGHT_BROWSERS_PATH: /ms-playwright` present under the harness job; the test-run
  step (PLAYWRIGHT_BASE_URL ~line 340) inherits it.
 - done: Playwright browsers persist across runs.
 ### Task 3 — Document the host-mount dependency
 Add a short subsection to `docs/DEVELOPMENT.md` (CI/runner area) noting:
 - CI now uses persistent caches at container paths `/pnpm-store` and `/ms-playwright`.
 - These require the act_runner `config.yaml` `container.options` to bind-mount host dirs to those
  paths (host change, not in this repo).
 - Without the mounts CI still works — it just falls back to uncached (writes to an ephemeral dir).
 - done: the host-side requirement is discoverable from the repo.
 ## Constraints
 - Must pass local gates before each commit: `format:check` (prettier), eslint, YAML validity
  (yq or actionlint if available), typecheck (no TS touched, but run if cheap).
 - Job names and the required-check contexts (CI / fast-checks, CI / api, CI / harness,
  CI / security, CI / gate) MUST stay identical so branch protection still matches. Do not rename
  jobs or restructure the job graph.
 - Atomic commits (Task 1, Task 2, Task 3 may be one or separate commits — keep changes coherent).
 ### Task 4 — Dockerfile BuildKit pnpm-store cache (added mid-task per user request)
 `apps/api/Dockerfile`: add `# syntax=docker/dockerfile:1` (line 1) and a BuildKit cache mount to
 all three pnpm install stages (builder, pwa-builder, production):
 `RUN --mount=type=cache,target=/pnpm-store,id=pnpm-store,sharing=locked pnpm install ... --store-dir /pnpm-store`.
 `sharing=locked` because builder + pwa-builder run in parallel and would otherwise race the store.
 `.gitea/workflows/publish.yml`: set `DOCKER_BUILDKIT: '1'` on the "Build production image" step so
 the legacy builder can't break on the `--mount` syntax (BuildKit is default on Docker 23+; explicit
 for safety).
 - done: image build reuses a persistent BuildKit pnpm-store cache across builds.
 ## OUT OF SCOPE (do not touch)
 - Verdaccio / any `.npmrc` registry change (deferred — user will set up later).
 ## must_haves
 - truth: "All four ci.yml pnpm installs use --store-dir /pnpm-store --prefer-offline"
 - truth: "The harness job sets PLAYWRIGHT_BROWSERS_PATH=/ms-playwright"
 - truth: "ci.yml remains valid YAML with unchanged job names / required-check contexts"
 - truth: "docs note the act_runner config.yaml host-mount requirement"
 - artifacts: [.gitea/workflows/ci.yml, docs/DEVELOPMENT.md]
@@ -0,0 +1,121 @@
 ---
 quick_id: 260618-tg2
 slug: persistent-ci-dependency-caches-pnpm-sto
 phase: "20"
 plan: tg2
 status: complete
 completed: 2026-06-18
 tags: [ci, caching, pnpm, playwright]
 key-files:
  modified:
    - .gitea/workflows/ci.yml
    - docs/DEVELOPMENT.md
    - apps/api/Dockerfile
    - .gitea/workflows/publish.yml
 decisions:
  - All four CI pnpm installs now target /pnpm-store via --store-dir --prefer-offline flags
  - PLAYWRIGHT_BROWSERS_PATH added at harness job level (not per-step) so both install and run steps share the same path
  - D-PROBE-04 comments updated to reflect the new store strategy rather than "no cache"
 ---
 # Quick Task 260618-tg2: Persistent CI dependency caches (pnpm store + Playwright) Summary
 **One-liner:** Point all four CI pnpm installs at `/pnpm-store` and harness Playwright at `/ms-playwright` via host-mounted directories on the act_runner.
 ## What Was Done
 ### Task 1 — pnpm store (ci.yml, 4 install lines)
 Changed all four `pnpm install --frozen-lockfile` lines to
 `pnpm install --frozen-lockfile --store-dir /pnpm-store --prefer-offline`:
 - Line ~54: `fast-checks` job
 - Line ~110: `api` job
 - Line ~197: `harness` job
 - Line ~485: `security` job (conditional)
 Both flags confirmed valid against `pnpm 11.5.1 install --help` before use.
 Updated the stale D-PROBE-04 comment in each job from "no cache backend / ~30s acceptable"
 to reflect that installs now target the host-mounted store.
 ### Task 2 — Playwright browsers (harness job env)
 Added `PLAYWRIGHT_BROWSERS_PATH: /ms-playwright` to the existing job-level `env:` block on
 the `harness:` job (alongside the DB_* creds). This means both the `Install Playwright browsers`
 step and the `Run harness` step inherit the same path, so cached binaries are found at install
 time and used at test time.
 Added a comment on the Playwright install step noting the future optimization: baking a runner
 image with browsers preinstalled would also eliminate the `--with-deps` apt step.
 ### Task 3 — Host-mount documentation (docs/DEVELOPMENT.md)
 Added a "CI dependency caches" subsection under the CI Pipeline Overview. Documents:
 - The two container paths (`/pnpm-store`, `/ms-playwright`) with a reference table
 - That the act\_runner `config.yaml` `container.options` bind-mount is a **host-side** change
 - That CI still works without the mounts (ephemeral fallback — just no caching)
 ## Verification
 ```
 grep -c -- '--store-dir /pnpm-store --prefer-offline' .gitea/workflows/ci.yml
 → 4
 grep -n 'PLAYWRIGHT_BROWSERS_PATH' .gitea/workflows/ci.yml
 → 187: PLAYWRIGHT_BROWSERS_PATH: /ms-playwright  (job-level env)
 → 278: # PLAYWRIGHT_BROWSERS_PATH=/ms-playwright... (comment)
 python3 -c 'import yaml,sys; yaml.safe_load(open(".gitea/workflows/ci.yml")); print("YAML valid")'
 → YAML valid
 pnpm format:check → All matched files use Prettier code style!
 pnpm md:lint     → Summary: 0 error(s)
 ```
 ### Task 4 — Dockerfile BuildKit pnpm-store cache (added mid-task by user request)
 Added by the orchestrator after the initial 3 tasks, when the user asked to include the Dockerfile:
 - `apps/api/Dockerfile`: added `# syntax=docker/dockerfile:1` (line 1) and a BuildKit cache mount
  (`RUN --mount=type=cache,target=/pnpm-store,id=pnpm-store,sharing=locked ... --store-dir /pnpm-store`)
  to all three pnpm install stages (builder, pwa-builder, production). `sharing=locked` because
  builder + pwa-builder run in parallel and would otherwise race the shared store.
 - `.gitea/workflows/publish.yml`: set `DOCKER_BUILDKIT: '1'` on the "Build production image" step —
  the publish path uses plain `docker build` (not buildx), and the legacy builder would fail on the
  `--mount` syntax. BuildKit is default on Docker 23+; set explicitly for safety.
 - Verified: `format:check` clean (Dockerfile is outside prettier's scope), publish.yml valid YAML.
 ## Commits
 | Hash | Message |
 | --- | --- |
 | `80b2038` | chore(20): persistent CI caches — pnpm store + Playwright browsers |
 | `f83d423` | docs(20): document CI persistent cache host-mount dependency |
 | `6e93e24` | chore(260618-tg2): BuildKit pnpm-store cache mount in Dockerfile build |
 ## Deviations from Plan
 Dockerfile cache (Task 4) was added mid-task at the user's request after the initial 3-task plan
 (it had been explicitly deferred/out-of-scope). The publish workflow's `DOCKER_BUILDKIT=1` was a
 required companion change so the `--mount` syntax doesn't break the legacy builder.
 - `--store-dir` flag form matches plan exactly (plan said verify against pnpm; verified: valid)
 - harness job already had a job-level `env:` map; `PLAYWRIGHT_BROWSERS_PATH` was added to it as instructed
 - stale comment text updated as instructed
 ## Known Stubs
 None.
 ## Threat Flags
 None — YAML-only and doc-only changes; no new network endpoints, auth paths, or trust boundaries introduced.
 ## Self-Check: PASSED
 - `.gitea/workflows/ci.yml` — modified and committed at 80b2038
 - `docs/DEVELOPMENT.md` — modified and committed at f83d423
 - YAML validity confirmed by python3 yaml.safe_load
 - 4 install lines confirmed by grep -c
 - PLAYWRIGHT_BROWSERS_PATH confirmed at job-level env line 187
@@ -11,7 +11,6 @@ FamilySync is a self-hosted, Dockerized family organization hub for a two-person
 ### Constraints
 - **Tech stack**: MariaDB for the database — PostgreSQL is not available in the stack
 - **Tech stack**: Redis available (optional, for live list sync / push)
 - **Infrastructure**: Unraid host running Docker + Docker Compose
 - **Auth**: Authelia (already deployed) + Local Auth — OIDC/OAuth2 for the custom app; all members authenticate through it
 - **Calendar backend**: Fastmail (paid, existing) is the single source for all calendars via JMAP/CalDAV
@@ -50,7 +49,6 @@ FamilySync is a self-hosted, Dockerized family organization hub for a two-person
 | web-push              | 3.6.7   | Server-side VAPID push           | Generate VAPID keys, sign and dispatch push messages to browser push services (APNs for iOS, FCM for Android)                                 |
 | @hono/oidc-auth       | 1.8.3   | OIDC session middleware for Hono | Storage-less JWT session cookies; authorization-code + PKCE flow; works with any RFC-compliant OIDC provider including Authelia               |
 | openid-client         | 6.8.4   | Low-level OIDC primitives        | If `@hono/oidc-auth` proves insufficient (e.g., custom token introspection), use this as the lower-level escape hatch                         |
 | ioredis               | 5.11.0  | Redis client                     | Pub/sub for broadcasting list-change events to SSE connections across Node processes                                                          |
 | zod                   | 3.24.x  | Schema validation                | Validate API request bodies and CalDAV event payloads before writing back to Fastmail                                                         |
 | @hono/zod-validator   | 0.8.0   | Hono middleware for Zod          | Validate request body/query in route handlers with Zod schemas                                                                                |
 | @tanstack/react-query | 5.101.0 | Server state + caching           | Manages calendar and list data fetching, background refetch, stale-while-revalidate; pairs with SSE for live list updates                     |
@@ -199,7 +197,7 @@ The backend handles two auth paths: local username/password (scrypt + HS256 JWT
 Calendar data lives exclusively in Fastmail CalDAV. The broker layer (`apps/api/src/broker/`) uses `tsdav` for PROPFIND/REPORT/PUT/DELETE, `ical.js` for VCALENDAR parsing, and `rrule` for server-side recurrence expansion. Writes are enqueued in a `calendarOutbox` table and drained asynchronously every 15 seconds; a ctag-based poller re-syncs calendars every 5 minutes.
-Lists are persisted in MariaDB. Live list updates flow over SSE (`text/event-stream`) via an in-process Node.js `EventEmitter`; a 30-second polling fallback is always active. Push notifications (reminders + calendar change alerts) are dispatched via `web-push` (VAPID) to APNs/FCM. Redis is present in the stack but not yet used at runtime (reserved for future multi-process pub/sub).
+Lists are persisted in MariaDB. Live list updates flow over SSE (`text/event-stream`) via an in-process Node.js `EventEmitter`; a 30-second polling fallback is always active. Push notifications (reminders + calendar change alerts) are dispatched via `web-push` (VAPID) to APNs/FCM.
 The PWA uses TanStack Query for all server state (events, lists, user, sync status, auth mode) and Zustand for UI-only state (selected date, open panels, active tab).
@@ -221,7 +219,7 @@ familysync/
 │       ├── hooks/           # useListSSE, usePushSubscription
 │       ├── store/           # Zustand stores (calendarStore, listsStore)
 │       └── sw.ts            # Custom Workbox service worker
-├── docker-compose.yml       # Production stack (api + mariadb + redis)
+├── docker-compose.yml       # Production stack (api + mariadb)
 └── docker-compose.dev.yml   # Dev overrides
 ```
@@ -15,7 +15,7 @@ A self-hosted family organization hub for a two-person household. One color-code
 - Node.js 22 LTS
 - pnpm 11.5.1 (`corepack enable pnpm`)
- Docker + Docker Compose (for MariaDB, Redis, and production deployment)
+- Docker + Docker Compose (for MariaDB and production deployment)
 ## Installation
@@ -53,7 +53,7 @@ Required environment variables (set in `.env` or your Docker host):
 ```bash
 # Start backing services
-docker compose -f docker-compose.yml -f docker-compose.dev.yml up mariadb redis
+docker compose -f docker-compose.yml -f docker-compose.dev.yml up mariadb
 # Run migrations
 pnpm --filter @familysync/api db:migrate
@@ -79,8 +79,8 @@ The API listens on port 3000. The PWA build is served separately (Vite `preview`
 apps/
  api/        Hono backend — CalDAV sync, OIDC auth, lists API, push notifications
  pwa/        React 19 PWA — calendar view, lists UI, service worker
-docker-compose.yml       Production services (API, MariaDB 11, Redis 7)
+docker-compose.yml       Production services (API, MariaDB 11)
-docker-compose.dev.yml   Dev overrides (bind-mount src/, expose DB/Redis ports)
+docker-compose.dev.yml   Dev overrides (bind-mount src/, expose DB ports)
 ```
 ## Commands
@@ -110,7 +110,7 @@ docker-compose.dev.yml   Dev overrides (bind-mount src/, expose DB/Redis ports)
 | Auth            | `@hono/oidc-auth` 1.8.3 — authorization code + PKCE against Authelia |
 | Calendar        | tsdav 2.2.2 (CalDAV) + ical.js 2.2.1 against Fastmail                |
 | Push            | web-push 3.6.7 (VAPID)                                               |
-| Live sync       | Server-Sent Events + Redis 7 pub/sub                                 |
+| Live sync       | Server-Sent Events (in-process EventEmitter)                         |
 | Frontend        | React 19, Vite 8, vite-plugin-pwa 1.3, TanStack Query 5, Zustand 5   |
 | Calendar UI     | Schedule-X 4.6                                                       |
@@ -1,3 +1,4 @@
 # syntax=docker/dockerfile:1
 # Built from the REPO ROOT context (see docker-compose.yml: build.context: .)
 # so the pnpm workspace manifest + lockfile are available for a deterministic,
 # workspace-aware install. apps/api is one package in the pnpm workspace.
@@ -12,7 +13,8 @@ FROM base AS builder
 COPY pnpm-workspace.yaml pnpm-lock.yaml package.json ./
 COPY apps/api/package.json ./apps/api/
 COPY apps/pwa/package.json ./apps/pwa/
-RUN pnpm install --frozen-lockfile --filter @familysync/api...
+RUN --mount=type=cache,target=/pnpm-store,id=pnpm-store,sharing=locked \
    pnpm install --frozen-lockfile --filter @familysync/api... --store-dir /pnpm-store
 COPY apps/api ./apps/api
 RUN pnpm --filter @familysync/api build
@@ -28,7 +30,8 @@ FROM base AS pwa-builder
 COPY pnpm-workspace.yaml pnpm-lock.yaml package.json ./
 COPY apps/api/package.json ./apps/api/
 COPY apps/pwa/package.json ./apps/pwa/
-RUN pnpm install --frozen-lockfile --filter @familysync/pwa...
+RUN --mount=type=cache,target=/pnpm-store,id=pnpm-store,sharing=locked \
    pnpm install --frozen-lockfile --filter @familysync/pwa... --store-dir /pnpm-store
 COPY apps/pwa ./apps/pwa
 RUN pnpm --filter @familysync/pwa build
@@ -36,7 +39,8 @@ FROM base AS production
 COPY pnpm-workspace.yaml pnpm-lock.yaml package.json ./
 COPY apps/api/package.json ./apps/api/
 COPY apps/pwa/package.json ./apps/pwa/
-RUN pnpm install --frozen-lockfile --prod --filter @familysync/api...
+RUN --mount=type=cache,target=/pnpm-store,id=pnpm-store,sharing=locked \
    pnpm install --frozen-lockfile --prod --filter @familysync/api... --store-dir /pnpm-store
 COPY --from=builder /app/apps/api/dist ./apps/api/dist
 WORKDIR /app/apps/api
 # Enforce production identity — engages the NODE_ENV=production hard guard
@@ -15,7 +15,6 @@ The stack must include:
 - API on `:3000` started with `DEV_AUTH_BYPASS=true` (see Security Guardrail below)
 - PWA dev server on `:5173` (`pnpm --filter @familysync/pwa dev`)
 - Dev MariaDB on `:3306` (exposed via `docker-compose.dev.yml`)
 - Redis on `:6379`
 **`DEV_AUTH_BYPASS=true` MUST be set in the API's environment BEFORE the API process starts.** The harness cannot inject it at runtime — the API reads the env var once at startup. If the API is running without it, all `/api/*` requests return an auth redirect and every spec fails.
@@ -5,7 +5,7 @@
 * Auth: DEV_AUTH_BYPASS=true on the API (never storageState — D-01/Pitfall 14)
 * SW: serviceWorkers: 'block' on all profiles (D-02/Pitfall 15)
 * baseURL: env-driven PLAYWRIGHT_BASE_URL (D-08/Rule 8)
- * webServer: manages Vite only — API+MariaDB+Redis stay compose-managed (D-10)
+ * webServer: manages Vite only — API+MariaDB stay compose-managed (D-10)
 *
 * Run:
 *   pnpm --filter @familysync/pwa test:e2e
@@ -63,7 +63,7 @@ export default defineConfig({
    },
  ],
-  // D-10: manage Vite only; API+MariaDB+Redis are compose-managed
+  // D-10: manage Vite only; API+MariaDB are compose-managed
  // reuseExistingServer: reuse operator's pnpm dev locally; start fresh in CI
  webServer: {
    command: 'pnpm --filter @familysync/pwa dev',
@@ -23,7 +23,3 @@ services:
  mariadb:
    ports:
      - '3306:3306'
  redis:
    ports:
      - '6379:6379'
@@ -50,9 +50,5 @@ services:
      timeout: 5s
      retries: 5
  redis:
    image: redis:7-alpine
    # Phase 1: present but unused; Phase 4 wires pub/sub for live list sync
 volumes:
  mariadb_data:
@@ -248,7 +248,6 @@ routes/setup.ts   ──→  db (app_config)
 | Local auth storage   | scrypt PHC hash in `local_credentials.password_hash`; session signed with `LOCAL_SESSION_SECRET` env var                    |
 | Push notifications   | `web-push` (VAPID) → APNs (iOS) / FCM (Android)                                                                             |
 | Live list sync       | In-process Node.js `EventEmitter` → SSE (`text/event-stream`)                                                               |
 | Redis                | Present in stack (image: `redis:7-alpine`); not used in current runtime (reserved for future multi-process pub/sub)         |
 | PWA                  | React 19 + Vite 8 + `vite-plugin-pwa` (Workbox `injectManifest` mode)                                                       |
 | Networking           | Pangolin/Newt tunnel — no open ports; split-DNS internal domain                                                             |
 | Deployment           | Docker Compose on Unraid; single `api` container serves both the API and the PWA static build                               |
@@ -163,7 +163,7 @@ VAPID_SUBJECT=mailto:admin@example.com
 ### Local Development (host-side)
-The dev Docker Compose override (`docker-compose.dev.yml`) exposes MariaDB on `localhost:3306` and Redis on `localhost:6379`. To run the API and PWA directly on the host:
+The dev Docker Compose override (`docker-compose.dev.yml`) exposes MariaDB on `localhost:3306`. To run the API and PWA directly on the host:
 ```bash
 # Build the API first (dev script runs compiled output)
@@ -35,7 +35,7 @@ src/
 - **Node.js 22 LTS** — the Dockerfile base is `node:22-alpine`; match this locally
 - **pnpm 11.5.1** — managed via corepack (`corepack enable pnpm`)
- **Docker + Docker Compose** — for MariaDB and Redis in dev
+- **Docker + Docker Compose** — for MariaDB in dev
 - **TypeScript 5.x** — installed per-workspace as a dev dependency
 ## Local Setup
@@ -48,13 +48,13 @@ pnpm install
 This installs all workspace packages (`apps/api` and `apps/pwa`) in a single pass.
-### 2. Start the dev database and Redis
+### 2. Start the dev database
 ```bash
-docker compose -f docker-compose.yml -f docker-compose.dev.yml up mariadb redis -d
+docker compose -f docker-compose.yml -f docker-compose.dev.yml up mariadb -d
 ```
-The dev override (`docker-compose.dev.yml`) exposes MariaDB on `localhost:3306` and Redis on `localhost:6379`.
+The dev override (`docker-compose.dev.yml`) exposes MariaDB on `localhost:3306`.
 ### 3. Configure environment variables
@@ -205,6 +205,21 @@ Every PR to `main` runs through `.gitea/workflows/ci.yml`. A `changes` path-filt
 All five jobs must pass (or be legitimately skipped) before a PR can merge. See [docs/TESTING.md](TESTING.md) for test suite details.
 ### CI dependency caches
 CI uses two persistent cache paths inside job containers:
 | Path             | Content                                                                 |
 | ---------------- | ----------------------------------------------------------------------- |
 | `/pnpm-store`    | pnpm content-addressable store (`--store-dir /pnpm-store`)              |
 | `/ms-playwright` | Playwright browser binaries (`PLAYWRIGHT_BROWSERS_PATH=/ms-playwright`) |
 These paths must be bind-mounted from host directories in the act_runner `config.yaml`
 `container.options` field — that is a **host-side change, not tracked in this repo**. Without the
 mounts, CI still works correctly — pnpm creates an ephemeral store at `/pnpm-store` inside the
 container and Playwright downloads browsers fresh each run. The mounts only eliminate repeat
 downloads across runs.
 ## Drizzle Migration Workflow
 Schema changes follow a strict two-step process. **`drizzle-kit push` is not available** — it has been removed from the scripts because it emits a false destructive diff (table truncation) on populated MariaDB databases.
@@ -238,17 +253,16 @@ Migration files live in `apps/api/src/db/migrations/` and are committed to versi
 ## Docker Compose Dev Stack
 ```bash
-# Bring up the full dev stack (API in Docker + MariaDB + Redis, with ports exposed)
+# Bring up the full dev stack (API in Docker + MariaDB, with ports exposed)
 docker compose -f docker-compose.yml -f docker-compose.dev.yml up
 # Bring up only backing services (run API on host for faster iteration)
-docker compose -f docker-compose.yml -f docker-compose.dev.yml up mariadb redis -d
+docker compose -f docker-compose.yml -f docker-compose.dev.yml up mariadb -d
 ```
 The dev override:
 - Exposes MariaDB on `localhost:3306`
 - Exposes Redis on `localhost:6379`
 - Mounts `apps/api/src` into the container for live source access
 - Sets `NODE_ENV=development`
@@ -12,7 +12,7 @@ This guide walks from a fresh clone to a running local development environment.
 | ----------------------- | ------------------ | ---------------------------------------------------------------------------------- |
 | Node.js                 | `22 LTS`           | Matches the `node:22-alpine` base in `apps/api/Dockerfile`                         |
 | pnpm                    | `11.5.1`           | Pinned in `package.json` `packageManager` field; enable via `corepack enable pnpm` |
-| Docker + Docker Compose | Any recent version | Used to run MariaDB and Redis locally                                              |
+| Docker + Docker Compose | Any recent version | Used to run MariaDB locally                                                        |
 **Node version management:** If you use nvm or fnm, install Node 22 LTS and set it as the default before continuing. There is no `.nvmrc` in the repo; the target version comes from the Dockerfile.
@@ -72,10 +72,10 @@ Open `.env` and fill in the required values. See [docs/CONFIGURATION.md](CONFIGU
 ### 5. Start the database services
 ```bash
-docker compose -f docker-compose.yml -f docker-compose.dev.yml up -d mariadb redis
+docker compose -f docker-compose.yml -f docker-compose.dev.yml up -d mariadb
 ```
-This starts MariaDB (bound to `localhost:3306`) and Redis (`localhost:6379`) using the dev override. Wait for MariaDB to pass its health check before proceeding.
+This starts MariaDB (bound to `localhost:3306`) using the dev override. Wait for MariaDB to pass its health check before proceeding.
 ### 6. Run database migrations
@@ -89,7 +89,7 @@ pnpm --filter @familysync/pwa test:e2e:ui
 pnpm --filter @familysync/pwa test:e2e:headed
 ```
-The `baseURL` is driven by `PLAYWRIGHT_BASE_URL` (default: `http://localhost:5173`). In local mode the config reuses a running Vite dev server; in CI it starts Vite itself. The API, MariaDB, and Redis must already be running via Docker Compose before launching e2e tests locally — see `docs/DEVELOPMENT.md`.
+The `baseURL` is driven by `PLAYWRIGHT_BASE_URL` (default: `http://localhost:5173`). In local mode the config reuses a running Vite dev server; in CI it starts Vite itself. The API and MariaDB must already be running via Docker Compose before launching e2e tests locally — see `docs/DEVELOPMENT.md`.
 ### Type checking (separate from tests — required)
@@ -16,11 +16,10 @@ Self-hosted Docker deployment on Unraid behind Authelia OIDC and a Pangolin/Newt
 The production compose file brings up three services:
-| Service   | Image                                                | Purpose                                                    |
+| Service   | Image                                                | Purpose                                             |
-| --------- | ---------------------------------------------------- | ---------------------------------------------------------- |
+| --------- | ---------------------------------------------------- | --------------------------------------------------- |
-| `api`     | Built from `apps/api/Dockerfile` target `production` | Hono API + compiled React PWA, listens on port 3000        |
+| `api`     | Built from `apps/api/Dockerfile` target `production` | Hono API + compiled React PWA, listens on port 3000 |
-| `mariadb` | `mariadb:11`                                         | Persistent MariaDB database                                |
+| `mariadb` | `mariadb:11`                                         | Persistent MariaDB database                         |
 | `redis`   | `redis:7-alpine`                                     | Present for live list sync (pub/sub); unused until Phase 4 |
 ---
Author	SHA1	Message	Date
Lucas BergerandClaude Opus 4.8	66e3b806be	chore: remove REQUIREMENTS.md for v1.1 milestone Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>	2026-06-18 22:04:01 -04:00
Lucas BergerandClaude Opus 4.8	7fbb3cca9d	chore: archive v1.1 milestone files Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>	2026-06-18 22:03:56 -04:00
luckberg	6cc3b8ae27	Merge pull request 'chore(ci): persistent pnpm store + Playwright caches (and Dockerfile BuildKit cache)' (#27 ) from gsd/quick-260618-tg2-ci-dep-cache into main Publish / publish (push) Successful in 1m23s Details Reviewed-on: #27	2026-06-18 21:32:06 -04:00
Lucas Berger	c5cdb9c21d	docs(quick-260618-tg2): persistent CI dependency caches (pnpm store + Playwright + Dockerfile) CI / changes (pull_request) Successful in 4s Details CI / api (pull_request) Successful in 2m6s Details CI / fast-checks (pull_request) Successful in 2m32s Details CI / security (pull_request) Successful in 1m2s Details CI / harness (pull_request) Successful in 5m36s Details CI / gate (pull_request) Successful in 2s Details	2026-06-18 21:21:26 -04:00
Lucas BergerandClaude Opus 4.8	6e93e24df0	chore(260618-tg2): BuildKit pnpm-store cache mount in Dockerfile build Add 'RUN --mount=type=cache,target=/pnpm-store' to all 3 pnpm install stages (builder/pwa-builder/production) with --store-dir /pnpm-store, plus the '# syntax=docker/dockerfile:1' directive. Set DOCKER_BUILDKIT=1 on the publish build step so the legacy builder can't break on the mount syntax. sharing=locked because builder and pwa-builder run in parallel. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>	2026-06-18 21:19:45 -04:00
Lucas Berger	f83d423e1c	docs(20): document CI persistent cache host-mount dependency - Add "CI dependency caches" subsection to CI Pipeline Overview - Lists /pnpm-store and /ms-playwright container paths - Notes act_runner config.yaml container.options host-mount requirement - Clarifies that CI still works without the mounts (ephemeral fallback)	2026-06-18 21:16:52 -04:00
Lucas Berger	80b20383f1	chore(20): persistent CI caches — pnpm store + Playwright browsers - All four pnpm install steps now use --store-dir /pnpm-store --prefer-offline - harness job env adds PLAYWRIGHT_BROWSERS_PATH: /ms-playwright - Updated stale D-PROBE-04 comments to reflect persistent store - Playwright install step gets a comment noting the future runner-image optimization	2026-06-18 21:16:43 -04:00
luckberg	2276a254e4	chore: remove unused Redis service and references (#26 ) Publish / publish (push) Successful in 23s Details	2026-06-18 21:06:32 -04:00
Lucas BergerandClaude Opus 4.8	0810260d0b	docs(20): add orphaned phase-20 UAT (7/7 passed, playwright-verified) UAT was committed locally after PR #25's final push, so it never reached main. Recovered and committed directly per maintainer authorization. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>	2026-06-18 20:48:52 -04:00