Convert the remaining 10 data sources (Bucket, Conversation, QuestionThread,
DatePlan, DateMemory, DateReflection, Answer, User, Auth, DateMatch) from
verbose suspendCancellableCoroutine{cont-> ...addOnSuccess/addOnFailure} wrappers
(and the per-file getDoc/voidAwait/queryAwait/refAwait helper duplicates) to the
kotlinx-coroutines-play-services .await() extension. The data layer now has ZERO
Task-callback wrappers.
Per-block review preserved non-trivial semantics: callbackFlow snapshot listeners
and the DateMatch transaction are untouched; failure→null/false best-effort blocks
became runCatching{ ...await() }.getOrNull()/getOrDefault() (e.g. markRevealed,
hasProfile, PlayIntegrity.verifyWithServer); no-signed-in-user guards became
`?: throw`.
Verified live (real Firebase, paired account): auth + Home reads (couple/user/
answer/challenge), daily-question answer WRITE (secure payload + metadata sets),
and the streak update all succeed — no crash, no PERMISSION_DENIED. Unit suite green.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Grounded review of the Android app / functions / build with measured
inconsistencies, what was standardized this session (collectAsStateWithLifecycle;
Firebase Tasks→await subset), and risk-assessed recommendations for the larger
items (finish Tasks→await, version catalog, ViewModel Firebase extraction,
Functions v2, color tokens, TS types).
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Replace verbose suspendCancellableCoroutine{ cont -> task.addOnSuccessListener
{cont.resume}.addOnFailureListener{cont.resumeWithException} } wrappers with the
kotlinx-coroutines-play-services .await() extension in the files whose blocks are
all one-shot Tasks (no snapshot listeners / transactions to endanger):
FirestoreOutcomeDataSource, FirestoreCoupleDataSource, PlayIntegrityChecker.
Notably fixes FirestoreOutcomeDataSource, which mixed both styles in one file.
PlayIntegrityChecker.verifyWithServer keeps its failure→null semantics via
runCatching{ ... .await() }.getOrNull() (a bare await would throw instead).
The remaining ~11 data sources also use the callback pattern but mix in
callbackFlow snapshot listeners / transactions, so they need per-block review —
tracked in Future.md rather than bulk-rewritten on untested I/O.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Migrate all 56 Compose screens from collectAsState() to the lifecycle-aware
collectAsStateWithLifecycle() (added androidx.lifecycle:lifecycle-runtime-compose).
Upstream flow collection now pauses when the UI isn't visible instead of running
in the background — the modern, recommended idiom, and it unifies the two files
already using it with the rest of the app.
No behavior change for visible screens; compiles + unit suite green.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Tracked dist output rebuilt to match src (aggregateOutcomeStats export +
aggregateOutcomes module), keeping the deployed bundle in sync with source.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Design for moving off the deprecated androidx.security:security-crypto to
Tink Android-Keystore-backed storage, without ever losing the couple key.
Covers: the load-bearing hazard (SecurePreferencesFactory.reset() silently
WIPES the couple key on any read failure) which must be removed first;
lazy dual-read + re-wrap + verify-then-clean migration; fail-closed (recover,
never delete) failure matrix; consumer ordering (couple key last); staged RC
rollout + content-free telemetry; test plan; open questions for the owner.
No implementation ships until the owner approves this design (Batch 5.1 gate).
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
aggregateOutcomeStats (scheduled) rolls up per-couple check-in deltas
(couples/{id}/outcomes/day_30|60|90) into the "X% feel closer in N weeks" stat
WITHOUT reading any E2EE content — only the self-reported deltas.
Privacy: counts/percentages only (no couple id or individual scores);
minimum-cohort suppression (N<50 → window omitted, no percentages); EXPORT-ONLY
via the top-level aggregate_stats collection, now explicitly deny-all in
firestore.rules (owner reads via console).
Pure aggregate()/extractCoupleOutcome() unit-tested incl. below-threshold
suppression + a no-PII assertion (6 tests); rules deny client read/write of
aggregate_stats (2 tests). Full functions 53/53, rules 121/121.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
- DateIdea gains sponsored/sponsorName/externalUrl (native, clearly-labeled ads
— no tracking SDK).
- FirestoreDateIdeaDataSource reads the server-curated date_ideas collection
(defensive: malformed docs skipped, https-only externalUrl); DateMatchRepo
serves remote first, falling back to the shipped static seed when empty/failed
so the feature always has content. Fallback verified live (Date Match renders
the seed) + unit-tested (remote-present / empty / failure).
- Date card shows a "Sponsored · <name>" pill and a "Learn more" link (opened via
the system browser, https-only, no click IDs).
- Local admin seeder (functions/scripts/seedDateIdeas.ts + date_ideas_seed.json,
generated from the Kotlin seed + 1 sponsored example) — idempotent by id, has a
--dry-run; NOT a deployed endpoint. Verified: 38 docs upserted to the emulator,
1 sponsored, tsc builds.
Owner: run the seeder against prod (GOOGLE_APPLICATION_CREDENTIALS) to populate
date_ideas; add real sponsor content.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
- PaywallPricing turns a RevenueCat Package into a headline price + an offer line
(free trial, intro price, or one-time), so a dashboard-configured trial or a
lifetime SKU renders without an app release; unknown shapes degrade to the
plain price. Pure formatting is unit-tested (PaywallPricingTest); the SDK
extraction is type-checked against purchases 8.20.
- Gift framing on the plan chooser ("One subscription covers you both — treat
your partner"): couple-shared premium already unlocks both partners, so this
is copy only, no billing change.
Live paywall (real offerings) still needs a real RC key + sandbox per
docs/SUBSCRIPTION_GO_LIVE.md (owner).
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Foreground notification taps carried only app_route, so their
push_notification_opened event logged the generic "app_route" category. Also
put the notification type on the intent so the event records the specific kind
(partner_answered, reveal_ready, ...). Backgrounded taps already carry type via
the FCM data payload. Completes push-opened coverage across both tap paths.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
The live couple streak used a raw 48h rule that reset on any 48h gap AND
inflated the count by incrementing on every answer (several a day counted
several times). Replace with StreakCalculator.nextCount: once per calendar day,
one forgiving grace day (a single missed day still continues), reset only after
2+ missed days; skip the redundant write on a same-day answer.
Celebration milestones now include the habit-building early tiers
(1/3/7/14/30/100/365), single-sourced from StreakCalculator so the count and the
celebration can't drift; the milestone dialog handles day 1 grammatically.
Tested (StreakCountTest): same-day idempotency, consecutive, grace, reset, and
partner timezone-skew (never breaks a same-boundary streak). Existing 24
StreakCalculator tests still pass.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Content-free daily-question state + streak, mirrored from HomeViewModel via a
dedicated DataStore (TodayWidgetUpdater -> TodayWidgetState). App-lock drops it
to a generic card so partner-activity state doesn't sit on the protected home
screen; signed-out/unpaired show neutral prompts. Tapping opens the daily
question via MainActivity's app_route extra.
Verified: provider registers, state bridge writes, headline logic unit-tested
(TodayWidgetTest), no crash on launch. On-home-screen render + a Remote Config
kill switch are tracked in Future.md.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
null