docs: apply V2 review fixes to roadmap.md — comprehensive update

P0 Critical: - Sync API: TBD → platform-service (Fastify, port 4003) - Cascade presets: aligned with PRD (Aggressive/Standard/Light/Minimal/None/Custom) - Deprecated APIs: NSUserNotificationCenter → UNUserNotificationCenter, ClockKit removed - Cloud sync: iCloud Phase 3-4, platform-service REST Phase 5 (no contradiction) - productId: chronomind registered in header + sync modules - Open-source: closed-source app, open timer engine lib - Design tokens: ChronoMind product section, own urgency colors P1 High: - Added pre-work (domain, wireframes, design tokens) - Added CI/CD (GitHub Actions + Vercel auto-deploy) - Added analytics (platform-service telemetry) - Added onboarding (3-step walkthrough) - Added privacy policy + terms of service - Added web accessibility (WCAG 2.1 AA) - Added timer accuracy tests + offline resilience strategy - Added feedback mechanism - Timer engine: clarified rAF (UI) + setInterval (SW) dual architecture - Neurodivergent mode: default UX, not toggle (Compact mode for power users) - Phase 3: macOS deferred to Phase 4, iPad layout added P2 Medium: - Calendar sync: two-way → one-way (read-only) - Event countdown timer added to Phase 2 - Timer export/import added to Phase 2 - iOS waitlist added to Phase 2 - Data export/deletion added to Phase 4 - NL extraction upgrade path added to Phase 4 - Body doubling deferred to v3 New appendices: Cost Estimation, Common Platform Reusability, Key Decisions Updated: success metrics (aligned with PRD), risk register (+4 new risks) V2 review doc marked as APPLIED (audit trail)
2026-02-27 20:32:25 -08:00 · 2026-02-27 20:32:25 -08:00 · 1b148757af
commit 1b148757af
parent c20bdc9b20
2 changed files with 218 additions and 53 deletions
--- a/docs/roadmap-v2-review.md
+++ b/docs/roadmap-v2-review.md
@ -2,8 +2,8 @@

 > **Purpose:** Systematic review of `roadmap.md` (v1) identifying bugs, gaps, missing details, and opportunities to leverage the `learning_ai_common_plat` shared ecosystem.
 > **Date:** February 2026
-> **Status:** PENDING USER REVIEW — do not implement until signed off.
-> **Action:** Review each finding, then this doc becomes the implementation guide for roadmap v2.
+> **Status:** ✅ APPLIED — All findings incorporated into `roadmap.md` v2.0. This doc is now an audit trail.
+> **Action:** No further action needed. See `roadmap.md` for the implementation-ready version.

 ---

--- a/docs/roadmap.md
+++ b/docs/roadmap.md
@ -1,9 +1,14 @@
 # ChronoMind — Product Roadmap

 > **Product:** ChronoMind — AI-Powered Time Awareness Layer
+> **Product ID:** `chronomind` (ByteLyst ecosystem)
 > **Strategy:** Option D — Web PWA (MVP) → SwiftUI (iOS/Watch/Mac) → Android (v2)
+> **Version:** 2.0 (revised per [roadmap-v2-review.md](roadmap-v2-review.md) audit)
 > **Created:** February 2026
-> **See also:** [PRD.md](PRD.md) | [INDUSTRY_RESEARCH.md](INDUSTRY_RESEARCH.md)
+> **License:** Closed-source app; timer engine library open-source (TBD license)
+> **See also:** [PRD.md](PRD.md) | [INDUSTRY_RESEARCH.md](INDUSTRY_RESEARCH.md) | [roadmap-v2-review.md](roadmap-v2-review.md)
+>
+> **Backend:** Platform-service in `learning_ai_common_plat` (Fastify, port 4003) — reuses 17+ existing modules (auth, subscriptions, stripe, invitations, referrals, settings, flags, telemetry, waitlist, ratelimit, usage, plans, licenses, etc.). Only 4 new modules needed: `timers`, `routines`, `shared-timers`, `households`.

 ---

@ -46,8 +51,8 @@ ChronoMind ships in **5 phases over ~6 months**, from web MVP to full cross-plat
 │  UI: Jetpack Compose                                                     │
 │  Watch: Wear OS + Glance widgets                                        │
 │                                                                          │
-│  Shared: Cloud Sync API (optional, E2E encrypted)                       │
-│  Analytics: Plausible / PostHog (privacy-first)                         │
+│  Shared: Platform-service (Fastify, port 4003) — Cosmos DB sync         │
+│  Analytics: Platform-service telemetry / Plausible (privacy-first)      │
 └──────────────────────────────────────────────────────────────────────────┘
 ```

@ -60,23 +65,35 @@ ChronoMind ships in **5 phases over ~6 months**, from web MVP to full cross-plat

 ### Week 1: Foundation + Timer Engine

+- [ ] **Pre-work (before coding)**
+  - [ ] Register domain: `chronomind.app` or `chronomind.io` (needed for deployment + Universal Links)
+  - [ ] Set up DNS for Vercel deployment
+  - [ ] Low-fi wireframes: timeline, create timer, Pomodoro, alarm overlay (1 day)
+  - [ ] Define urgency level color mapping from `@bytelyst/design-tokens` + ChronoMind product section
+
 - [ ] **Project scaffold**
  - [ ] `npx create-next-app@latest` with Next.js 16, App Router, TypeScript
  - [ ] TailwindCSS v4 setup + shadcn/ui init + Lucide React icons
  - [ ] Zustand store with IndexedDB persistence (via `idb`)
  - [ ] ESLint + Prettier + Vitest config
  - [ ] Basic folder structure: `app/`, `components/`, `lib/`, `styles/`
+  - [ ] GitHub Actions CI: lint + typecheck + vitest on PR
+  - [ ] Auto-deploy to Vercel on `main` push
+  - [ ] Analytics: platform-service telemetry module (or Plausible as lightweight fallback)
+  - [ ] Track: page views, timer created, timer completed, cascade fired, Pomodoro completed, PWA installed

 - [ ] **Timer engine (`lib/timer-engine.ts`)**
  - [ ] `Timer` interface with all fields (id, type, label, urgency, targetTime, duration, cascade, etc.)
  - [ ] Create timer (alarm, countdown, pomodoro)
  - [ ] Start / pause / resume / cancel countdown
  - [ ] Timer state machine: `idle → active → warning → firing → snoozed → dismissed`
-  - [ ] `requestAnimationFrame` + `setInterval` dual-loop for accurate countdown
+  - [ ] `requestAnimationFrame` for UI countdown display (smooth 60fps in active tab)
+  - [ ] `setInterval` in Service Worker for background notification scheduling (30s check)
+  - [ ] Both contexts needed: rAF stops when tab hidden, SW runs in background
  - [ ] Unit tests for timer state transitions

 - [ ] **Pre-warning cascade (`lib/cascade.ts`)**
-  - [ ] Cascade presets: Relaxed (2h, 1h, 30m), Standard (1h, 30m, 15m, 5m), Tight (30m, 15m, 5m, 1m), Critical (2h, 1h, 30m, 15m, 5m, 1m)
+  - [ ] Cascade presets (aligned with PRD): Aggressive (4h, 3h, 2h, 90m, 1h, 30m, 15m, 5m, 1m), Standard (2h, 1h, 30m, 15m, 5m), Light (1h, 15m, 5m), Minimal (15m), None (fire only), Custom (user-defined)
  - [ ] Custom cascade builder (user picks intervals)
  - [ ] Cascade scheduler: calculate all warning timestamps from target time
  - [ ] Cascade state tracking: which warnings have fired
@ -172,6 +189,34 @@ ChronoMind ships in **5 phases over ~6 months**, from web MVP to full cross-plat
  - [ ] "Try it now" button → app dashboard (no signup needed)
  - [ ] Social proof section (placeholder for future testimonials)
  - [ ] Footer: links, GitHub, privacy
+  - [ ] Privacy policy page (`/privacy`) — required before any data collection or App Store submission
+  - [ ] Terms of service page (`/terms`)
+
+- [ ] **First-run onboarding**
+  - [ ] 3-step walkthrough: create timer → set cascade → see timeline
+  - [ ] "Create your first timer" prompt on empty dashboard
+  - [ ] Tooltip hints on cascade and urgency pickers
+
+- [ ] **Feedback mechanism**
+  - [ ] Feedback button in app (opens email or form)
+  - [ ] Bug report button with auto-attached browser info
+
+- [ ] **Web accessibility (WCAG 2.1 AA)**
+  - [ ] Keyboard navigation works for all flows
+  - [ ] Screen reader tested (NVDA/VoiceOver) on timeline and timer creation
+  - [ ] Focus indicators on all interactive elements
+  - [ ] Sufficient color contrast for all urgency levels
+
+- [ ] **Timer accuracy tests**
+  - [ ] Timing accuracy: create timer for T+5s, verify fires within 100ms tolerance
+  - [ ] Service Worker notification timing tests
+  - [ ] Stress test: 50 concurrent timers with cascades → all fire correctly
+
+- [ ] **Offline resilience strategy**
+  - [ ] Document SW lifecycle: timers survive SW termination via IndexedDB timestamps + SW `periodicsync`
+  - [ ] Background sync queue for analytics events
+  - [ ] IndexedDB size limit handling (warn user, purge old history)
+  - [ ] "Tab must be open for notifications" disclaimer for iOS Safari

 ### Phase 1 Exit Criteria

@ -183,8 +228,12 @@ ChronoMind ships in **5 phases over ~6 months**, from web MVP to full cross-plat
 - [ ] Lighthouse PWA score > 90
 - [ ] Page load < 2 seconds
 - [ ] All Vitest unit tests pass
- [ ] Deployed to Vercel/Netlify with custom domain
- [ ] 5 beta testers using it daily
+- [ ] Deployed to Vercel with custom domain
+- [ ] CI/CD pipeline running (GitHub Actions + Vercel auto-deploy)
+- [ ] Analytics tracking timer creation and cascade engagement
+- [ ] WCAG 2.1 AA: keyboard nav + screen reader tested
+- [ ] Privacy policy published at `/privacy`
+- [ ] 5 beta testers using it daily (measured via analytics)

 ---

@ -287,7 +336,7 @@ ChronoMind ships in **5 phases over ~6 months**, from web MVP to full cross-plat
  - [ ] Conflict detection: warn if imported event overlaps existing timer
  - [ ] Map calendar event priority to urgency level

- [ ] **Neurodivergent mode**
+- [ ] **Neurodivergent-friendly design (DEFAULT UX, not a toggle)**
  - [ ] Visual countdown ring (`components/CountdownRing.tsx`)
    - [ ] Large, animated circular countdown (inspired by Time Timer)
    - [ ] Color transitions: green → yellow → orange → red as time decreases
@ -295,8 +344,8 @@ ChronoMind ships in **5 phases over ~6 months**, from web MVP to full cross-plat
  - [ ] Gentle transition alerts (soft chime + visual pulse, not jarring alarm)
  - [ ] Reduced cognitive load: hide advanced options by default, progressive disclosure
  - [ ] Time blindness aids: "This is about as long as [familiar reference]" (e.g., "25 min — one TV episode")
-  - [ ] Body doubling placeholder: "Focus with others" room (v2 feature, show coming soon)
-  - [ ] Toggle in settings: "Gentle mode" (affects sounds, animations, notification style)
+  - [ ] Body doubling placeholder: "Focus with others" room (v3 feature, show coming soon)
+  - [ ] Toggle in settings: "Compact mode" for power users who want dense UI (the DEFAULT is already neurodivergent-friendly per PRD decision #7)

 - [ ] **Prep time intelligence**
  - [ ] Per-timer: "I need X minutes to prepare"
@ -310,6 +359,22 @@ ChronoMind ships in **5 phases over ~6 months**, from web MVP to full cross-plat
  - [ ] Suggestion shown as dismissable card on dashboard
  - [ ] Accept → auto-adjusts recurring timer

+- [ ] **Event countdown timer**
+  - [ ] Days/hours until a future date with milestone warnings (30 days, 7 days, 1 day, etc.)
+  - [ ] Timer type `event` in data model (matching PRD feature #26)
+  - [ ] Visual: large countdown display with milestone progress bar
+  - [ ] Use cases: "132 days until wedding", "47 days until vacation"
+
+- [ ] **Timer export / import (local backup)**
+  - [ ] "Export timers" → JSON download of all timers, routines, stats
+  - [ ] "Import timers" → upload JSON to restore
+  - [ ] Warning: "Your timers are stored locally. Export regularly or enable cloud sync (coming in v1.1)."
+
+- [ ] **iOS waitlist**
+  - [ ] Waitlist signup on web landing page ("Get notified when iOS launches")
+  - [ ] Use platform-service `waitlist` module (already exists) when backend is connected
+  - [ ] Fallback: simple email collection form
+
 - [ ] **Timer history (`app/(app)/history/`)**
  - [ ] Searchable log of all past timers
  - [ ] Filter by category, urgency, date range
@ -331,18 +396,19 @@ ChronoMind ships in **5 phases over ~6 months**, from web MVP to full cross-plat
 - [ ] Focus mode blocks notifications during session
 - [ ] Streak tracking persists across sessions
 - [ ] Calendar .ics import works with Google Calendar exports
- [ ] Neurodivergent mode noticeably gentler than default
+- [ ] Neurodivergent mode is default and noticeably gentler than previous
 - [ ] All E2E tests pass in CI
 - [ ] 25+ beta testers, 3+ using routines daily
- [ ] App Store / ProductHunt submission prepared
+- [ ] ProductHunt submission prepared (web) — App Store prep moved to Phase 3

 ---

 ## Phase 3: Apple Native — "Own the Wrist" (Week 6-9)

-> **Goal:** Ship native iOS + Apple Watch + macOS apps that make ChronoMind the best timer experience on Apple devices.
+> **Goal:** Ship native iOS + Apple Watch apps. macOS deferred to Phase 4 (4 weeks for iOS + Watch + Live Activities is already aggressive).
 > **Success criteria:** Apple Watch complications are the #1 reason users love ChronoMind.
 > **Framework decision:** Fully native SwiftUI (NOT KMP) — see [platform analysis in PRD](PRD.md).
+> **Note:** This phase is 4 weeks, not 3 as PRD states. PRD phases section should be updated to match.

 ### Week 6: iOS App Foundation

@ -436,15 +502,9 @@ ChronoMind ships in **5 phases over ~6 months**, from web MVP to full cross-plat
  - [ ] App Intents framework for iOS 17+
  - [ ] Shortcuts app integration

-### Week 8: macOS + AI Reschedule
+### Week 8: AI Reschedule + Gamification

- [ ] **macOS menu bar app (`ChronoMindMac/`)**
-  - [ ] Menu bar icon: shows countdown of next timer
-  - [ ] Click → popover with mini timeline
-  - [ ] Quick timer creation from popover
-  - [ ] Notification center integration (native macOS notifications)
-  - [ ] Keyboard shortcut: `Cmd+Shift+T` → quick timer
-  - [ ] Share 90%+ code with iOS via Shared/ folder
+> **Note:** macOS menu bar app deferred to Phase 4 (Week 10-11). Phase 3 focuses on iOS + Watch quality.

 - [ ] **AI reschedule**
  - [ ] "I slept in 30 minutes" → shift all morning timers by 30m
@ -459,7 +519,12 @@ ChronoMind ships in **5 phases over ~6 months**, from web MVP to full cross-plat
  - [ ] Weekly summary card (shareable image — key for viral growth)
  - [ ] Gentle celebration animations (confetti for streak milestones, not obnoxious)

-### Week 9: Polish + TestFlight
+### Week 9: iPad + Polish + TestFlight
+
+- [ ] **iPad layout**
+  - [ ] iPad sidebar navigation (NavigationSplitView)
+  - [ ] Wider timeline with side panel for timer detail
+  - [ ] Stage Manager / multiple window support (iPadOS 16+)

 - [ ] **Cross-device continuity**
  - [ ] Handoff: start timer on Mac, see it on iPhone and Watch
@ -483,34 +548,44 @@ ChronoMind ships in **5 phases over ~6 months**, from web MVP to full cross-plat
  - [ ] External beta: 100 testers via TestFlight link
  - [ ] Crash reporting: built-in MetricKit
  - [ ] Feedback form in Settings
+  - [ ] App Store submission prepared (screenshots, description, keywords, preview video)

 ### Phase 3 Exit Criteria

 - [ ] iOS app installs and all features work on iPhone 14+ (iOS 17)
+- [ ] iPad layout works with sidebar navigation (iPadOS 16+)
 - [ ] Apple Watch complications update correctly on Watch Series 8+ (watchOS 10)
 - [ ] Live Activities show on Dynamic Island and Lock Screen
 - [ ] Widgets update on home screen and lock screen
- [ ] macOS menu bar app works on macOS 14+
 - [ ] Siri Shortcuts work for top 3 use cases
 - [ ] AI reschedule correctly shifts timers
 - [ ] All XCTests and XCUITests pass
 - [ ] TestFlight beta: 100 external testers, crash-free rate > 99%
- [ ] App Store submission prepared (screenshots, description, keywords)
+- [ ] App Store submission prepared and submitted (screenshots, description, keywords, preview video)
+- [ ] macOS deferred to Phase 4

 ---

 ## Phase 4: Context & Growth — "Make It Sticky" (Week 10-14)

-> **Goal:** Calendar sync, location awareness, sleep integration, and viral growth mechanics.
+> **Goal:** macOS menu bar, calendar sync, location awareness, cloud sync, and viral growth mechanics.
 > **Success criteria:** Users say "ChronoMind knows my schedule better than I do."

-### Week 10-11: Calendar + Location
+### Week 10-11: macOS + Calendar + Location
+
+- [ ] **macOS menu bar app (`ChronoMindMac/`)** — deferred from Phase 3
+  - [ ] Menu bar icon: shows countdown of next timer
+  - [ ] Click → popover with mini timeline
+  - [ ] Quick timer creation from popover
+  - [ ] Notification center integration (UNUserNotificationCenter)
+  - [ ] Keyboard shortcut: `Cmd+Shift+T` → quick timer
+  - [ ] Share 90%+ code with iOS via Shared/ folder

 - [ ] **Full calendar sync**
  - [ ] Google Calendar OAuth 2.0 integration
  - [ ] Apple Calendar (EventKit) read access
  - [ ] Outlook Calendar (Microsoft Graph API) integration
-  - [ ] Two-way sync: imported events become timers with auto-cascade
+  - [ ] One-way sync: imported events become timers with auto-cascade (read-only — ChronoMind does NOT write back to calendars)
  - [ ] Conflict resolution: calendar events take priority, manual timers preserved
  - [ ] Sync frequency: every 15 minutes + on-demand refresh
  - [ ] Visual distinction: synced events vs manual timers on timeline
@ -549,14 +624,27 @@ ChronoMind ships in **5 phases over ~6 months**, from web MVP to full cross-plat
  - [ ] Optional: prompt at consistent times (morning, post-lunch, evening)
  - [ ] All data local — never sent to any server

- [ ] **Cloud sync (optional, privacy-first)**
-  - [ ] E2E encrypted sync via CloudKit (Apple) or custom API
+- [ ] **Cloud sync (Apple-only via iCloud, privacy-first)**
+  - [ ] iCloud/CloudKit sync for Apple devices only (iPhone ↔ Mac ↔ Watch)
  - [ ] Sync scope: timers, routines, categories, preferences, stats
  - [ ] Conflict resolution: last-write-wins with vector clocks
  - [ ] Opt-in: works perfectly offline, sync is a bonus
-  - [ ] Web: sync via authenticated API (JWT + encrypted payload)
+  - [ ] **Cross-platform sync (web ↔ Android) deferred to Phase 5** via platform-service REST API
+  - [ ] TLS in transit + encrypted at rest in Cosmos DB (E2E encryption deferred — timer data is not sensitive enough to justify complexity)
  - [ ] Delete account: wipes all cloud data immediately

+- [ ] **Data export / account deletion (App Store + GDPR requirement)**
+  - [ ] "Export all my data" → JSON download of all timers, routines, stats
+  - [ ] "Delete my account" → wipe all cloud data, revoke tokens
+  - [ ] Required before App Store submission (Apple requirement since 2022)
+
+- [ ] **Advanced NL parsing (upgrade path from chrono-node)**
+  - [ ] Define `timer-parse` extraction task in extraction-service (port 4005)
+  - [ ] Classes: `time`, `duration`, `urgency_hint`, `label`, `recurrence`
+  - [ ] Use `@bytelyst/extraction` client for LLM-powered parsing
+  - [ ] Falls back to chrono-node if extraction-service unavailable
+  - [ ] Enables: "Set a critical alarm for my dentist next Tuesday at 2pm with 1 hour warning and 30 min travel time"
+
 ### Week 13-14: Viral Growth Mechanics

 - [ ] **Shareable weekly summary**
@ -607,10 +695,12 @@ ChronoMind ships in **5 phases over ~6 months**, from web MVP to full cross-plat

 ### Phase 4 Exit Criteria

- [ ] Calendar sync works with Google, Apple, Outlook calendars
+- [ ] macOS menu bar app works on macOS 14+ (deferred from Phase 3)
+- [ ] Calendar sync works with Google, Apple, Outlook calendars (one-way)
 - [ ] Location triggers fire reliably (iOS geofencing)
 - [ ] Travel time estimates adjust pre-warning cascade
 - [ ] Sleep data integration working via HealthKit
+- [ ] Data export and account deletion functional
 - [ ] Weekly summary card shareable to 3+ social platforms
 - [ ] Shareable timer links work (web + iOS universal link)
 - [ ] ProductHunt page live, launch scheduled
@ -680,13 +770,14 @@ ChronoMind ships in **5 phases over ~6 months**, from web MVP to full cross-plat
  - [ ] Haptic pre-warnings
  - [ ] Dismiss/snooze from wrist

- [ ] **Cross-platform cloud sync**
-  - [ ] Sync API: REST endpoints for timer CRUD + sync
-  - [ ] JWT authentication (reuse from web)
-  - [ ] E2E encryption: client-side encrypt before upload
+- [ ] **Cross-platform cloud sync (via platform-service)**
+  - [ ] Sync API: platform-service REST endpoints for timer CRUD + sync (`productId: "chronomind"`)
+  - [ ] JWT authentication via @bytelyst/auth (reuse existing module)
+  - [ ] TLS in transit + encrypted at rest in Cosmos DB
  - [ ] Sync protocol: last-write-wins with conflict detection
-  - [ ] Works across: Web ↔ iOS ↔ macOS ↔ Android
-  - [ ] Account creation: email + password or Apple/Google Sign-In
+  - [ ] Works across: Web ↔ iOS ↔ macOS ↔ Android (replaces Phase 4 iCloud-only sync)
+  - [ ] Account creation: Apple Sign-In / Google Sign-In (email/password requires platform-service Email Delivery + Password Reset — may not be ready)
+  - [ ] Register `chronomind` as product in platform-service products module

 ### Week 20-21: Shared Timers + Social

@ -699,12 +790,9 @@ ChronoMind ships in **5 phases over ~6 months**, from web MVP to full cross-plat
  - [ ] Household admin: manage members, remove shared timers
  - [ ] Up to 6 members per household (Family tier)

- [ ] **Body doubling / shared focus (experimental)**
-  - [ ] "Focus room" — join a virtual co-working session
-  - [ ] See other participants' focus timers (anonymized)
-  - [ ] Accountability: "3 people are focusing with you"
-  - [ ] Optional: audio ambiance shared in room
-  - [ ] Join public rooms or create private (link invite)
+- [ ] **Body doubling / shared focus (deferred to v3)**
+  - [ ] _Experimental feature requiring real-time infrastructure (Event Bus). Defer until platform gap component 2.4 is built._
+  - [ ] v3 scope: "Focus room" — virtual co-working, see others' timers (anonymized), shared ambiance

 - [ ] **API + webhooks**
  - [ ] REST API: CRUD timers, routines, categories
@ -768,10 +856,13 @@ ChronoMind ships in **5 phases over ~6 months**, from web MVP to full cross-plat

 ### Metrics That Matter

-| Metric | MVP Target | v1 Target | v2 Target |
-|--------|-----------|-----------|-----------|
-| **MAU** | 50 | 1,000 | 5,000 |
-| **DAU/MAU ratio** | > 40% | > 50% | > 50% |
+| Metric | MVP Target (Phase 1-2) | v1 Target (Phase 3-4) | v2 Target (Phase 5) |
+|--------|------------------------|----------------------|---------------------|
+| **PWA installs** | 100 | 500 | 2,000 |
+| **DAU** | 50 | 200 | 1,000 |
+| **MAU** | 150 | 1,000 | 5,000 |
+| **DAU/MAU ratio** | > 33% | > 50% | > 50% |
+| **Timers created/week** | 500+ | 5,000+ | 25,000+ |
 | **Pre-warning usage** | > 60% of timers | > 70% | > 75% |
 | **Routine creation** | > 20% of users | > 30% | > 35% |
 | **7-day retention** | > 30% | > 40% | > 50% |
@ -780,6 +871,7 @@ ChronoMind ships in **5 phases over ~6 months**, from web MVP to full cross-plat
 | **Pro conversion** | — | > 3% | > 5% |
 | **Streak > 7 days** | — | > 15% of users | > 25% |
 | **Shared summary** | — | > 10% weekly | > 20% |
+| **Snooze rate** | < 40% | < 30% | < 25% |

 ### Viral Growth Loops

@ -851,8 +943,12 @@ Family ($5.99/mo):       Everything in Pro + shared timers (6 members),
 | 6 | Low App Store visibility | Medium | 3-4 | ASO from day 1; ProductHunt launch; content marketing |
 | 7 | Sync conflicts across platforms | Medium | 4-5 | Start with iCloud (Apple-only); cross-platform sync in Phase 5 |
 | 8 | Android notification restrictions (Doze, exact alarms) | Medium | 5 | Use `AlarmManager.setExactAndAllowWhileIdle()`; request permissions |
-| 9 | User overwhelm — too many features | Medium | 2+ | Progressive disclosure; smart defaults; "simple mode" toggle |
+| 9 | User overwhelm — too many features | Medium | 2+ | Progressive disclosure; smart defaults; neurodivergent-friendly default |
 | 10 | Monetization too early kills growth | Low | 4 | Free for 3 phases; only monetize when PMF is proven |
+| 11 | Platform-service Email Delivery not ready for Phase 4 accounts | Medium | 4 | Fallback: Apple Sign-In / Google Sign-In only (skip email/password) |
+| 12 | Platform-service Event Bus not ready for Phase 5 shared timers | Medium | 5 | Fallback: polling (30s interval) instead of real-time push |
+| 13 | Service Worker termination loses timer state | Medium | 1 | Persist all state to IndexedDB; SW checks timestamps on wake, not in-memory |
+| 14 | Design token conflict with ByteLyst ecosystem | Low | 1 | ChronoMind gets own product section in token file; shared base, own semantics |

 ---

@ -877,8 +973,77 @@ Family ($5.99/mo):       Everything in Pro + shared timers (6 members),
 |----------|-----------|--------------|---------|---------------|
 | **Web** | Next.js 16, React 19 | TailwindCSS v4, shadcn/ui, Zustand, Serwist, chrono-node, date-fns, Recharts | IndexedDB (idb) | Web Push + Service Worker |
 | **iOS** | SwiftUI (iOS 17+) | SwiftData, WidgetKit, ActivityKit, EventKit, CoreLocation | SwiftData | UNUserNotificationCenter |
-| **watchOS** | SwiftUI (watchOS 10+) | WidgetKit (complications), ClockKit | App Groups shared | WKExtension notifications |
-| **macOS** | SwiftUI (macOS 14+) | AppKit (menu bar) | SwiftData (shared via iCloud) | NSUserNotificationCenter |
+| **watchOS** | SwiftUI (watchOS 10+) | WidgetKit (complications) | App Groups shared | WKExtension notifications |
+| **macOS** | SwiftUI (macOS 14+) | AppKit (menu bar) | SwiftData (shared via iCloud) | UNUserNotificationCenter |
 | **Android** | Jetpack Compose | Material 3, Room, Hilt, Glance (widgets), AlarmManager | Room (SQLite) | NotificationManager + AlarmManager |
 | **Wear OS** | Compose for Wear OS | Tiles, Complications, Rotary Input | Room (shared) | Wearable notifications |
-| **Sync API** | TBD (Fastify or Cloudflare Workers) | JWT auth, E2E encryption | Cosmos DB or D1 | — |
+| **Sync API** | Platform-service (Fastify 5, port 4003) in `learning_ai_common_plat` | @bytelyst/auth (JWT), @bytelyst/cosmos, @bytelyst/config | Azure Cosmos DB (`productId: "chronomind"`) | Platform-service notifications module |
+
+---
+
+## Appendix B: Cost Estimation
+
+| Item | Phase | Monthly Cost | Notes |
+|------|-------|-------------|-------|
+| **Vercel (web hosting)** | 1+ | $0 (free tier) | Hobby plan covers MVP easily |
+| **Domain** | 1+ | ~$12/yr | `.app` or `.io` domain |
+| **Azure Cosmos DB** | 4+ | ~$25-50/mo | Serverless tier, `productId: "chronomind"` partition |
+| **Apple Developer Program** | 3+ | $99/yr | Required for App Store + TestFlight |
+| **Google Play Console** | 5+ | $25 one-time | Required for Play Store |
+| **Plausible Analytics** | 1+ | $0 (self-hosted) or $9/mo | Or use platform-service telemetry (free) |
+| **Stripe** | 5+ | 2.9% + 30¢ per transaction | Pro/Family subscription processing |
+| **Total (Phase 1-3)** | | ~$10/mo | Minimal — mostly free tiers |
+| **Total (Phase 4-5)** | | ~$75-100/mo | Cosmos DB + domain + Apple program |
+
+> **Note:** Platform-service infrastructure (Cosmos DB, AKV, health checks) is shared with LysnrAI. ChronoMind's incremental cost is only additional Cosmos RUs for timer data.
+
+---
+
+## Appendix C: Common Platform Reusability
+
+ChronoMind reuses **13 shared packages** and **17+ platform-service modules** from `learning_ai_common_plat`:
+
+| Package | ChronoMind Use | Phase |
+|---------|---------------|-------|
+| @bytelyst/auth | JWT auth for Pro/Family accounts | 4-5 |
+| @bytelyst/react-auth | Web dashboard auth context | 4-5 |
+| @bytelyst/api-client | Web → sync API calls | 4-5 |
+| @bytelyst/config | Config + `productId: "chronomind"` + AKV | 4-5 |
+| @bytelyst/cosmos | Cosmos DB for cloud sync, shared timers | 4-5 |
+| @bytelyst/errors | Typed HTTP errors | 4-5 |
+| @bytelyst/logger | Server-side structured logging | 4-5 |
+| @bytelyst/design-tokens | Cross-platform design system | 1+ |
+| @bytelyst/monitoring | Health-check utilities | 5 |
+| @bytelyst/blob | Custom alarm sound uploads | 5 |
+| @bytelyst/extraction | Advanced NL parsing via extraction-service | 4+ |
+| @bytelyst/testing | Shared test mocks | 4-5 |
+| @bytelyst/fastify-core | Service app factory | 5 |
+
+**New platform-service modules needed (4):**
+- `timers` — timer CRUD + sync for `productId: "chronomind"`
+- `routines` — routine CRUD + sync
+- `shared-timers` — household shared timer management
+- `households` — household membership + invitations
+
+---
+
+## Appendix D: Key Decisions
+
+| # | Decision | Options | Recommendation | Rationale |
+|---|----------|---------|----------------|-----------|
+| 1 | Backend stack | Platform-service vs Cloudflare Workers | **Platform-service** | Reuses 17+ existing modules |
+| 2 | Apple-only sync | CloudKit vs skip until Phase 5 | **CloudKit for Phase 3-4** | Free, fast for Apple-only |
+| 3 | E2E encryption | Full E2E vs TLS + at-rest | **TLS + at-rest** | Timer data not sensitive enough |
+| 4 | Real-time shared timers | WebSocket vs SSE vs polling | **SSE + polling fallback** | WebSocket overkill for timers |
+| 5 | Design tokens | Extend bytelyst.tokens.json | **ChronoMind product section** | Shared base, own urgency colors |
+| 6 | Analytics | Plausible vs PostHog vs telemetry | **Platform-service telemetry** | Free, self-hosted, integrated |
+| 7 | Phase 3 scope | iOS + Watch + Mac vs iOS + Watch | **iOS + Watch; Mac in Phase 4** | 4 weeks already aggressive |
+| 8 | Calendar sync | Two-way vs one-way | **One-way (read-only)** | Two-way is fragile |
+| 9 | Body doubling | Phase 5 vs v3 | **Defer to v3** | Needs Event Bus infrastructure |
+| 10 | Domain name | chronomind.app vs .io | **Check availability pre-Phase 1** | Needed for deployment |
+| 11 | Open-source | Full vs closed vs hybrid | **Hybrid: open timer engine lib** | Credibility + community |
+| 12 | Windows/Tauri | Phase 5 vs Phase 6 vs v3 | **Defer to v3** | Focus on mobile first |
+| 13 | AKV secrets | Own CHRONO_SECRETS vs share | **Share Cosmos/JWT, own Stripe** | productId partitions data |
+| 14 | Fonts | ByteLyst (DM Sans) vs own (Inter) | **Own: Inter + JetBrains Mono** | Superior digit rendering |
+
+> **Windows/Tauri note:** PRD lists Windows as v2.0 target. Deferred to v3 — insufficient market signal to justify Tauri investment before mobile platforms are stable. Revisit after Phase 5 Android launch.