[1.0.0.4] — 2026-04-28
Planned focus
Usability polish, ecosystem expansion, and higher-confidence automation workflows.
Implemented so far
- Optimized System health checks:
- Parallelize integration probes instead of checking services one by one
- Cache health probe results briefly so the System page and Health Matrix do not duplicate network checks
- Refresh the Integration Status panel on an interval to match the Health Matrix freshness
- Added production frontend chunk splitting for charts, sockets, icons, and remaining vendor code to reduce oversized main bundle risk
- Added route-level lazy loading so major pages are loaded on demand instead of bundled into the initial app payload
- Added automation preflight checks before full-cycle start:
- Blocks cycles when critical services/search sources are unavailable
- Warns about queue saturation, pending failed-download cleanup, optional integration issues, and low disk headroom
- Adds a System page preflight panel with per-check status details
- Improved Queue page UX:
- Unified active/recent refresh path with manual refresh and 15-second fallback polling
- Added active/completed/failed summary counters
- Added recent-download status filters and timestamps
- Improved responsive row layout for long release names and error messages
- Added service-health cache invalidation after settings saves so connection status reflects configuration changes quickly
- Improved Settings page review workflow:
- Added inline validation for malformed URLs, invalid numeric ranges, incomplete active downloader config, and missing search sources
- Added quick-jump section navigation for long Settings pages on smaller windows
- Added downloader capability display for SABnzbd/NZBGet support coverage
- Added a download-client capability matrix and API so future clients can declare support for submit, queue/history status, purge, categories, pause/resume, and storage-path lookup
- Added active-downloader capability checks to automation preflight
- Added installer frontend asset-manifest smoke test so packaging fails early if
frontend/dist/index.htmlreferences missing built assets - Added additive SQLite startup migrations for existing installs so failed-download retry/cleanup metadata columns are created without resetting user data
- Hardened SABnzbd failure recovery:
- Capture SAB failure messages and storage paths for incomplete/aborted jobs
- Purge failed jobs from both SAB queue and history
- Clean failed download folders automatically before retrying alternatives
- Route manual and scheduled downloads through a shared monitor/replace/cleanup/retry workflow
- Improved retry behavior:
- Retry only accepted replacement candidates
- Skip every release already attempted for the movie
- Avoid retrying generic blacklisted releases even if they came from a different indexer
- Mark movies as failed when all retry/replacement paths are exhausted instead of leaving them stuck as downloading
- Improved orphan cleanup:
- Match downloader jobs against both raw and client-prefixed Slimarr job IDs
- Capture SAB orphan storage paths correctly
- Make the Orphaned Downloads cleanup action purge downloader history and delete orphaned files/folders immediately
UI and usability (planned)
- Sidebar resilience on smaller windows:
- Ensure navigation remains reachable when viewport height is constrained
- Keep footer actions visible while allowing menu items to scroll
- Additional responsive polish pass across pages with long action stacks beyond System/Settings/Queue
Integration opportunities under investigation
- Bazarr companion integration:
- Trigger subtitle refresh/search after successful replacements
- Surface missing subtitle counts per media item in Slimarr UI
- Lidarr integration (music):
- Evaluate extending Slimarr's scoring/replacement model to albums/tracks
- Reuse existing downloader/indexer plumbing for audio workflows
- Whisparr parity integration:
- Reuse Sonarr-like API patterns for monitor/unmonitor and post-import sync
- Keep feature toggled/optional like current Radarr/Sonarr integration approach
- Readarr status note:
- Readarr is currently retired by the Servarr team, so direct integration is lower priority unless ecosystem support stabilizes
Reliability and automation candidates
- Smarter retry windows (time-based backoff + indexer/uploader failure weighting)
- Integration health history (not just latest state) for trend-based diagnostics
- Persist preflight and health snapshots for trend views instead of keeping only the latest result
Packaging and performance candidates
- Installer/package footprint audit (exclude unnecessary test modules from packaged runtime where safe)