Loading ToolBench
arr-mcp demonstrates solid naming conventions (verb_noun pattern: bazarr_subtitles, arr_orchestrate, lidarr_artists) and comprehensive parameter schemas with enums for constrained inputs. However, descriptions vary significantly in quality, some are detailed and context-rich (arr_orchestrate, bazarr_subtitles), while others are generic (arr_health: 'Health-check the *arr stack'). Output schemas are not documented in the visible code. Error handling guidance is absent. Tool composition is reasonable but some tools conflate multiple concerns (arr_orchestrate combines media server checks + auto-routing). Parameter descriptions are present but inconsistent in depth.
Show upcoming releases as a rich Prefab calendar card. Aggregates upcoming releases from Radarr, Sonarr, Lidarr, and Readarr into a single visual card.
Health-check the *arr stack — probes system status per service.
Show the *arr stack health as a rich Prefab card. Returns a PrefabApp card showing reachability and version for every configured *arr service. Equivalent to `arr_health(service='all')` but rendered as a visual card in supporting MCP clients.
Discover and learn about arr-mcp tools. **discover** — list all registered tools with their operations. **tool_info** — get detailed documentation for a specific tool. **quickstart** — step-by-step guide to start using arr-mcp.
Cross-arr orchestration: check Jellyfin/Plex/Emby, auto-route to correct arr, stack status. **THE DIFFERENTIATOR** — chains media server availability checks (Jellyfin + Plex + Emby) with automatic *arr routing. "I want to watch Dune" → checks configured media servers → if not found → auto-adds to Radarr.
Output schemas not documented. Tools return data but LLMs cannot predict field structure for downstream chaining. E.g., arr_orchestrate returns media metadata but fields are not specified.
Generic descriptions lack context for tool selection. arr_health ('Health-check the *arr stack'), arr_health_card, arr_stats_card provide minimal guidance on when/why to call them vs alternatives.
No error recovery guidance. Tools lack descriptions of failure modes, retryability, or next steps. E.g., bazarr_subtitles 'download' operation does not document what happens if provider fails or subtitle not found.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | D | 56 | 2026-07-28+ | v2 |
Show consolidated stack statistics as a rich Prefab card. Returns a PrefabApp card with media counts and wanted/missing per service.
Manage Bazarr subtitles: wanted list, search, download, history, providers. Bazarr is the subtitle automation companion for the *arr stack. Results from this tool can feed jellyfin-mcp's RAG pipeline for downstream subtitle quality analysis.
Manage Lidarr albums: list, get, lookup, set monitored.
Manage Lidarr artists: list, search, add, delete, update.
Manage Overseerr media requests: view, create, approve, decline, delete. Overseerr is the user-facing request gateway for the *arr stack. Users request movies/TV shows through Overseerr, which routes to Radarr/Sonarr for fulfillment.
Search Overseerr for movies, TV shows, and people across TMDB.
Manage Overseerr users and their request history.
Tool composition concern: arr_orchestrate combines media server availability checks + auto-routing to *arr services. Should split into separate tools (check_media_servers, request_media) for independent composition.
Prefab card tools (arr_health_card, arr_calendar_card, arr_stats_card) lack input/output schema documentation. arr_calendar_card accepts 'days' but no bounds specified; return format (PrefabApp structure) not documented.