bun's mock.module persists across test files in the same process,
causing image.test.ts to receive the stub instead of the real
tmdbImageUrl. Since tmdbImageUrl has no heavy runtime dependencies,
service tests work fine with the real implementation.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- 211 tests across 15 files covering services (tracking, discovery,
metadata, backup, credits, webhooks, settings, update-check, colors,
person), utilities (config, cron, providers, title-theme), and TMDB
image URL helpers
- `lib/test-preload.ts` + `bunfig.toml` wire up a global in-memory
SQLite DB with migrations for all DB-backed tests
- `lib/test-utils.ts` provides `clearAllTables()` and seed helpers
(insertUser, insertTitle, insertTvShow, insertMovieWatch, etc.)
- Export `getBackupSource`, `isKnownBackup`, `isValidBackupFilename`,
`buildBackupCron`, `performUpdateCheck` internals for direct testing
- GitHub Actions workflow runs `bun test --coverage` on push/PR to main
- Backup and update check jobs show as "disabled" (not "succeeded") when
their respective settings are off
- Disabled jobs excluded from "x of y jobs healthy" count, sorted to
bottom of list, with schedule/next-run hidden and trigger button disabled
- Active jobs sorted by next run time
- "Never run" uses amber dot to differentiate from gray "disabled" dot
- Extract RefreshButton component and add it to all three health cards
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Remove the static package.json import in lib/version.ts and instead
pass APP_VERSION as a Docker build arg extracted from package.json by
the CI workflow. Both the builder and runner stages receive APP_VERSION
and GIT_COMMIT_SHA as environment variables, making runtime version
info available via process.env without bundling package.json into the
standalone output.
Build a provider registry mapping TMDB provider IDs to search URL
templates. `generateProviderUrl()` resolves a URL for a given provider
and title name using URL-encoded search queries. `readAvailability()`
now accepts the title name and attaches `watchUrl` to each offer.
Provider badges with a resolved URL render as `<a>` links opening in a
new tab; tooltip copy switches from the provider name to "Watch on
{name}". Badges without a known URL remain non-interactive as before.
When a show had status "watchlist", logging episode watches did not
transition it to "in_progress", so it never appeared in the Continue
Watching feed which filters for in_progress titles only.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Cast, availability, recommendations, colors, and trailers were missing
on first load because importTitle fire-and-forgot enrichment tasks.
getTitleWithChildren now calls ensureEnriched() to backfill any missing
data before rendering — works for new imports, old titles, and shells.
Also removes the awaitEnrichment option from importTitle (redundant now)
and optimizes DB queries: reuses already-read data for existence checks,
caches fetchSeasonsFromDb result, and skips re-reads when no backfill ran.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Store TMDB genres in normalized tables (genres + titleGenres) and content
ratings as a column on titles. Both are fetched during import/refresh using
append_to_response for content ratings (no extra API calls). Displayed in
the title hero between type badge and year. Also fixes pre-existing type
error for profile_path on TmdbSearchResult.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Add composite indexes on userEpisodeWatches and userMovieWatches for hot queries
- Batch episode tracking: wrap season/batch watches in single transaction (~8 queries vs 8*N)
- Fix N+1 patterns in credits, recommendations, and filmography with batch prefetch+insert
- Stream TV season hydration via Suspense instead of blocking page render
- Optimize webhook logs (per-connection LIMIT 10) and system health queries
- Merge genre filter waterfalls into single Promise.all fetch
- Migrate deprecated Jotai loadable() to unwrap()
- Replace isolated createStore()+Provider with useHydrateAtoms on root store
- Remove unnecessary atomWithStorage SSR guards (handled by Jotai internally)
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Coalesces concurrent imports via an in-flight promise map, and catches
SQLITE_CONSTRAINT_UNIQUE as a safety net to return the existing row.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Motion's initial="hidden" rendered opacity:0 in server HTML, making content
invisible until JS hydrated — causing a multi-second blank gap after skeleton
disappearance. Replaced with CSS @keyframes stagger animation that plays
immediately from SSR. Also made color extraction fire-and-forget, merged
duplicate stats queries, and removed redundant dashboard loading.tsx.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Integrate TMDB credits data into the app: cast carousels on title detail
pages, person detail pages with biography and filmography, and person
search results in the command palette. Includes new persons/titleCast
DB tables, profile image caching, and a nightly credits refresh cron job.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Fetch trailer video keys from TMDB videos API during title import/refresh,
store in DB, and display a "Trailer" button on title pages that opens a
media-chrome Sutro-themed YouTube player in a fullwidth dialog.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Add try/catch around `request.json()` in all POST routes to return a
400 instead of crashing on malformed bodies. Validate `type` as a
strict `"movie" | "tv"` enum in search, discover, import, and resolve
routes. Validate `tmdbId` as a positive integer. Add a
`SORT_BY_PATTERN` regex and page-range check (1–500) to the discover
route. Wrap all outbound TMDB calls in try/catch and return 502 on
failure so clients get a structured error rather than an unhandled
rejection.
Fix optimistic-update rollbacks in `use-title-actions`: capture
`prevStatus` and `prevWatches` before each mutation and restore both
atoms in the catch block for catchUp, handleMarkSeason,
handleUnmarkSeason, and single-episode toggle.
Fix a bug in `getContinueWatchingFeed` where the watchDateMap could
hold a stale date for episodes watched more than once; the map now
keeps the most-recent `watchedAt` per episode.
Serialize all backup operations through a promise-based queue
(`withBackupLock`) to prevent race conditions under concurrent
requests. Extract internal helpers so `pruneBackups` and
`deleteBackup` can participate in the same lock.
Strengthen `restoreFromBackup`: write the temp file alongside the
live database (`dbDir`), validate it with a full integrity check,
foreign key check, and a required-tables list before touching the
live DB, then use a synchronous `renameSync` for an atomic swap.
Clear WAL/SHM files synchronously in the same tick to avoid
interleaving. Run migrations after restore so older backups are
brought up-to-date automatically.
Add milliseconds to all backup timestamps (`HHmmssSSS`) to avoid
filename collisions when backups are created in rapid succession.
Update filename regexes to accept both the old and new formats.
Add createLogger to tmdb client, image-cache, availability, and
webhooks services. Replace all silent .catch(() => {}) in metadata.ts
with debug-level error logging. Add progress and success logs in cron
job bodies, colors extraction, and webhook processing. Fix pre-existing
unused parameter lint warning in title-card.tsx.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Replace manual fetch/useState/useEffect patterns with SWR hooks in
system health section and command palette search, gaining automatic
caching, request deduplication, and stale-while-revalidate.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Replace nested loop queries with batch fetches using inArray() across
discovery, tracking, metadata, settings, and system-health services.
The biggest win is getContinueWatchingFeed() going from ~375+ queries
to 5. Add a cached getSession() wrapper via React.cache() to deduplicate
auth lookups shared between layouts and pages. Parallelize independent
async operations in importTitle(), cacheImagesJob(), and the explore page.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Rebuild the background jobs card as a sortable table showing each job's
schedule, last run time (live-updating via a new useTimeAgo hook), last
duration, and a manual trigger button backed by a new POST
/api/admin/jobs/trigger route. Extract StatusDot into a shared
component. Add cronToHuman() to display schedule patterns as readable
strings (e.g. "Every 6h", "Daily at 03:00"). Replace static
formatDistanceToNow calls throughout the health section with a
LiveTimeAgo component that refreshes every 30 seconds. Also swap a
handful of section icons for better visual matches across settings cards.
Checks current version against latest GitHub release every 6 hours via
cron job, caches result in appSettings, and surfaces updates through a
toast notification (once per browser session) and an animated badge in
the settings footer. Includes an admin toggle to enable/disable checks
(enabled by default). Also renames ServerSection to RegistrationSection
now that update checks are in their own card.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Shows a subtle progress bar at the bottom of title cards indicating
watched/total episodes, with a tooltip for exact counts. Uses a single
efficient SQL query with JOINs to batch-fetch progress for all visible
titles on the Explore page.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Title cards now show the user's existing watch status (watchlist,
watching, completed) with a color-coded badge on the poster and a
status-aware quick-add button that prevents re-adding tracked titles.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Replace ad-hoc useState/useEffect fetch patterns with Jotai atoms and
loadables in StatsDisplay, FilterableTitleRow, BackupScheduleSection,
IntegrationsSection, and CommandPalette. Each component now gets a
scoped Jotai Provider with a pre-initialized store so server-rendered
initial values hydrate correctly. Async data fetching moves into
atom-level loadables, eliminating manual loading flags and cancellation
logic throughout.
Add Emby as a third media server integration alongside Plex and Jellyfin.
Emby webhooks use a similar payload format (JSON with nested Item object
and ProviderIds). Also removes the mediaServerUsername field from all
webhook connections since it was never used for authentication — the
token-in-URL is the sole auth mechanism. This simplifies the connection
UX to a single "Connect" button.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Move all server actions from scattered page-level files into shared lib/actions/
directory (settings.ts, titles.ts, watchlist.ts) so they can be reused across
the app. Add a hover-triggered plus button on explore page title cards that lets
users add titles to their watchlist without navigating away.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Add admin-only system health API and UI showing database stats, TMDB
connection status, background job history, storage usage, and environment
variables (with redacted secrets). Split monolithic backup card into
three focused cards (backups, schedule, restore) with dedicated section
headers for Server, Security, and Backups.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Add "unlimited" (value 0) to the keeping-last dropdown
- Skip pruning when retention is set to unlimited
- Show toast confirmation when changing retention
- Update description to read "Keeping unlimited backups" or "Keeping last N backups"
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Make backup frequency (6h/12h/1d/7d), time, and day-of-week
configurable via settings, stored in appSettings key-value table
- Add rescheduleBackup() to dynamically update cron job at runtime
- Differentiate manual (sofa-manual-*) vs scheduled (sofa-scheduled-*)
backup filenames with source icons and tooltips in the backup list
- Pruning now only targets scheduled backups, leaving manual ones intact
- Merge backup-actions.ts into actions.ts, use shadcn DropdownMenu for
all selectors (consistent with dashboard stats cards)
- Fix hydration mismatch on relative backup timestamps
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Created a new `cronRuns` table to track job execution history, including job name, status, timestamps, and error messages.
- Updated job scheduling logic to log job status in the `cronRuns` table, capturing success and error states.
- Enhanced error handling to log detailed error messages in the database.
This commit establishes a foundation for monitoring scheduled jobs and their outcomes.
Background sparkline area charts on Movies and Episodes cards show watch
activity distribution over the selected period. Periods now use rolling
windows (past 24h/7d/30d/365d) instead of calendar boundaries. Bumped
stats grid breakpoint to lg. Upgraded recharts 2.15.4 → 3.7.0 and fixed
chart.tsx types for v3 (TooltipContentProps, LegendPayload).
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Movies and Episodes cards now have an interactive dropdown to switch
between Today, This Week, This Month, and This Year. Extracts reusable
getWatchCount helper in discovery service and adds /api/stats route.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Backup service using VACUUM INTO for WAL-safe atomic snapshots
- Server-side storage in DATA_DIR/backups with download/upload API routes
- Restore with integrity validation and automatic pre-restore safety backup
- Scheduled daily backups via cron with configurable retention
- Admin-only settings UI consolidated under single Server section
- Clean up account section sign-out button, fix switch sub-pixel rendering
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Replaces the 335-line React Context provider with a scoped Jotai store,
giving consumers granular subscriptions and stable handler callbacks.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
drizzle-orm/bun-sqlite is fully synchronous — all queries return values
directly, not promises. Remove await from all db calls, drop async from
functions that no longer need it, simplify Promise.all patterns that
wrapped sync operations, and fix setSetting() which was missing .run()
(previously masked by await triggering execution via thenable).
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Swap the hand-rolled Scheduler class (setInterval-based) for croner,
a battle-tested cron library with overlap protection and proper cron
expressions. Consolidate lib/jobs/ into a single lib/cron.ts and
remove the unused admin jobs API route.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Override CSS custom properties (--primary, --ring, --primary-foreground,
--status-watching) on the title detail page wrapper using the vibrant
color from the poster's node-vibrant palette. All Tailwind classes
referencing these variables automatically pick up the new color.
Also await lazy color extraction in getTitleWithChildren() so palettes
are available on first visit to never-before-opened titles.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Status (watchlist/watching/completed) for TV shows is now automatically
determined by episode progress rather than requiring manual selection from
a dropdown. The only manual action is adding/removing from watchlist —
watching and completed states flow from episode data.
- Decouple markAllEpisodesWatched from setTitleStatus and export it
- Replace status dropdown with toggle button + read-only status badge
- Add "Mark All Watched" button with confirmation dialog for TV shows
- Fix optimistic transitions (watchlist → watching on first episode)
- Simplify W keyboard shortcut to toggle watchlist on/off
- Add optimistic completion check when marking entire seasons
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Replace @libsql/client with bun:sqlite for zero-dependency SQLite access,
swap drizzle-orm/libsql adapter for drizzle-orm/bun-sqlite, and rewrite
Dockerfile to use oven/bun:1-alpine. The raw Database instance is no longer
exported via Proxy (native C++ methods lose `this` binding through
Reflect.get); instead, a closeDatabase() helper handles graceful shutdown
and the health check uses Drizzle's db.run() directly.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Replace separate DATABASE_URL and IMAGE_CACHE_DIR env vars with a
single DATA_DIR root (default: ./data, Docker: /data). DATABASE_URL
and CACHE_DIR are derived from it but can still be overridden
individually.
Also:
- Pin pnpm to @10 for reproducible Docker builds
- Add --link to COPY instructions for better BuildKit cache reuse
- Add syntax directive for BuildKit features
- Simplify Dockerfile mkdir to just /data (ensureImageDirs handles subdirs)
- Remove redundant DATABASE_URL from docker-compose.yml
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Support self-hosted OIDC providers (Authentik, Authelia, Keycloak, etc.)
configured entirely via environment variables. Uses Better Auth's
hooks.before to gate email/password sign-up at the endpoint level, and
disables emailAndPassword entirely when DISABLE_PASSWORD_LOGIN is set.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
The DB client was initialized at import time, which caused Next.js page
data collection to fail during Docker builds where no database exists.
Wrapping in Proxy defers createClient() until first actual use.
Also adds data/ to .dockerignore.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Remove react-hotkeys-hook, KeyboardProvider context, useRegisterShortcut
wrapper, and KeyboardHelpDialog component. Components now call TanStack
useHotkey/useHotkeySequence directly. Shared state uses jotai atoms in
lib/atoms/, shortcut metadata lives as a static const in lib/constants/.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Replace fully client-side dashboard and title detail pages with server
component orchestrators that fetch data directly via service functions,
eliminating extra network round-trips through API routes. Each section
streams independently through its own Suspense boundary.
- Extract getUserStats(), getTitleWithChildren(), getRecommendationsForTitle()
into service layer; update getNewAvailableFeed() to include tmdbId/voteAverage
- Split dashboard into server sections (stats, continue watching, library,
recommendations) with client children for animations
- Split title page into server hero + client interaction provider with shared
context for optimistic mutations across actions and seasons
- Add generateMetadata with OG tags, server-side TMDB ID resolution via
redirect(), loading.tsx and not-found.tsx for both pages
- Add per-section skeleton components, fix ContinueWatchingSkeleton dimensions
- Remove 6 unused API routes (feed/*, titles/[id] GET, titles/[id]/recommendations)
- Remove components/stats-summary.tsx, add lib/types/title.ts for shared types
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
jakeandClaude Opus 4.6