Skip to content

Prowlarr And Indexer Sources

Gadarr can search through Prowlarr or through direct Torznab/Newznab/RSS-compatible sources. Prowlarr remains the preferred Arr-ecosystem aggregator, but it is not required for Gadarr to operate.

Configuration

Set these server environment variables:

PROWLARR_URL=http://127.0.0.1:9696
PROWLARR_API_KEY=...
PROWLARR_GAME_CATEGORIES=4000,4050

PROWLARR_GAME_CATEGORIES is optional. If omitted, Gadarr uses the default game-oriented category set 1000,4000,4050 to avoid broad TV/movie searches. Set it to all or none to search without forcing a category filter. Settings -> Indexers shows categories discovered from GET /api/v1/indexer/categories as selectable pills for saved Torznab/Newznab/RSS-compatible sources, while the built-in Prowlarr source remains environment-managed.

When these variables are present, Gadarr exposes Prowlarr as a virtual source in Settings -> Indexers. It is not stored in SQLite and can be changed by editing environment configuration.

Settings -> Indexers also exposes Sync From Prowlarr. That action reads Prowlarr's enabled indexers, keeps only torrent or Usenet indexers with search or RSS support, intersects their advertised categories with Gadarr's configured game categories, and creates or updates saved Gadarr Torznab/Newznab sources that point to Prowlarr's per-indexer /{id}/api proxy endpoints.

Direct Source Configuration

Users can add saved sources in Settings -> Indexers:

  • torznab: searches a Torznab/Newznab-compatible XML endpoint with t=search, q, cat, apikey, and limit.
  • newznab: same request pattern, mapped as a Usenet-capable source when returned items are not torrents.
  • rss: polls a plain feed URL and filters feed items against watched game titles and aliases.
  • fanzub: builds Fanzub RSS/search URLs with anime defaults and maps results as Usenet releases.
  • nyaa: builds Nyaa RSS/search URLs with anime category defaults and maps seed/leech metadata.
  • iptorrents and torrentleech: poll tracker RSS feeds, filter feed items locally, and keep automatic/interactive search disabled by default.
  • broadcasthenet: calls BroadcastheNet's JSON-RPC getTorrents API with the saved API key and maps torrent metadata.
  • filelist: calls FileList's JSON API with username/passkey basic auth and maps download/detail URLs.
  • hdbits: posts HDBits JSON API torrent queries with username/API key auth and maps torrent metadata.

Saved sources support independent toggles for manual search, automatic wanted search, and RSS sync. Categories, source priority, and minimum seeders are kept server-side. Torznab/Newznab sources can be queried for t=caps category metadata through native source routes and Servarr-compatible category actions.

Background RSS and automatic searches skip enabled saved sources that are in a short failure backoff after three consecutive errors. Interactive searches still allow the source so recovery can be tested immediately.

Auth

Prowlarr expects the API key in the X-Api-Key header. Torznab/Newznab sources usually expect apikey as a query parameter. Gadarr sends either only from the API service and never returns stored keys to the browser.

Discovery Flow

  1. GET /api/v1/system/status to identify the Prowlarr server.
  2. GET /api/v1/health to surface blocking health issues.
  3. GET /api/v1/indexer to pull enabled indexers, protocol, capabilities, privacy, priority, and status.
  4. GET /api/v1/indexer/categories to let users choose game-related categories.
  5. POST /api/sources/prowlarr/sync to materialize matching Prowlarr indexers as saved Gadarr game indexer sources.

Search Flow

  1. Gadarr receives a watched game from a root-folder scan or manual entry.
  2. Gadarr selects enabled sources for manual or automatic search.
  3. Gadarr searches with the game title and configured category ids when present.
  4. For Prowlarr, Gadarr calls GET /api/v1/search with:
  5. query
  6. repeated indexerIds when scoped to selected indexers
  7. repeated categories when category filters are configured
  8. limit
  9. offset
  10. For direct Torznab/Newznab and RSS-compatible named indexers, Gadarr fetches the XML feed and maps RSS items plus torznab:attr, newznab:attr, and named-feed metadata into the same release shape. API-backed named indexers use their JSON APIs and map results into the same release shape.
  11. Release decision logic compares titles against the game title and aliases.
  12. Gadarr records accepted/rejected candidates and search history before anything reaches the queue.

Prowlarr's source maps this endpoint through a Newznab request and may cache returned releases for its own grab operations. Gadarr records returned releases as candidates, applies its own game release classifier and profiles, then owns the queue record and protocol-aware download-client handoff itself.

RSS Sync Flow

POST /api/rss/sync and the RSS Sync system task run source-level polling:

  1. Select sources with RSS enabled.
  2. For plain RSS feeds and named RSS-compatible sources, fetch the feed once and match items to every watched game by title and aliases.
  3. For Prowlarr/Torznab/Newznab sources, perform per-game searches so category and seeder rules still apply.
  4. Save release candidates and search history.
  5. Auto-grab the best acceptable monitored candidate when no active queue or existing queue record already covers it.
  6. Update source last-sync fields and failure counters.

RSS sync follows the same safety rules as manual grabs: rejected, blocklisted, DRM-bypass, missing-download, and unsafe DLC-only candidates are not sent to the download client.

The release-filter UI starts from a game-first view: base-game-style rows are shown first, unversioned/unknown rows are hidden by default, and repack/crack toggles are explicit review controls. Crack/keygen/DRM-bypass releases remain blocked from grabs even when a user includes them for inspection.

Prowlarr Sync Option

Prowlarr sync is category-gated on purpose. If Gadarr is configured with game categories and a Prowlarr indexer advertises only TV, movie, music, or book categories, Gadarr skips it instead of adding a broad source that could surface unrelated releases. Saved synced sources preserve Prowlarr priority, RSS/search flags, tags, and the Prowlarr API key server-side; the browser only receives redacted source configuration.

Direct Prowlarr search remains available through the virtual Prowlarr source, and users who do not run Prowlarr can keep using direct Torznab, Newznab, RSS, and named tracker sources.