Prowlarr

Prowlarr

Bindery does not have a dedicated Prowlarr indexer type — and you don't need one. Prowlarr exposes every indexer it tracks through a Newznab-compatible endpoint (for Usenet indexers) or a Torznab endpoint (for torrent trackers). Add those to Bindery as normal Newznab / Torznab indexers and everything just works.

This page documents the exact setup.

---

Usenet indexers (Newznab)

In Prowlarr:

1. Add and test the Usenet indexer normally (Prowlarr → Indexers → Add Indexer).
2. Click the indexer and note its Indexer URL. It looks like:

   http://prowlarr.local:9696/1/api?apikey=<prowlarr-api-key>
   

   — the number after the host is Prowlarr's internal indexer ID.
3. Grab the Prowlarr API key from Settings → General → Security.

In Bindery:

1. Settings → Indexers → Add → Newznab.

2. Fill in:

   Field	Value
   Name	Anything (e.g. Prowlarr — NZBGeek)
   URL	http://prowlarr.local:9696/1 (drop the ?apikey=…)
   API Key	The Prowlarr API key
   Categories	7020 (ebooks), 3030 (audiobooks) — Prowlarr preserves category IDs

3. Click Test. A green check means Prowlarr forwarded the query and the upstream indexer answered.

Repeat once per indexer you want exposed to Bindery. There is no advantage to bundling multiple trackers behind a single Prowlarr entry — Bindery already fan-outs queries in parallel, so adding them individually keeps per-indexer error visibility.

---

Torrent trackers (Torznab)

Identical flow, except:

1. In Bindery choose Settings → Indexers → Add → Torznab.
2. The URL path uses the same http://prowlarr.local:9696/<indexer-id> form.
3. Pair it with a Torznab-capable download client (qBittorrent, Deluge, Transmission) — Bindery grabs the .torrent, hands it to the client, and the client does the actual transfer.

---

Troubleshooting

• 401 on Test. The API key is the Prowlarr key, not the upstream indexer's key. Prowlarr owns the auth and forwards using its own credentials to the tracker.
• Empty results on a known-good query. Make sure the category IDs match what Prowlarr exposes. Some trackers remap IDs; Prowlarr shows the effective ID on its indexer detail page.
• Test passes but no grabs happen. Check Prowlarr's History → Grabs tab. If grabs show there but the download never starts, the issue is between Prowlarr and your download client, not Bindery.
• Connection refused from inside a container. Use the Prowlarr service DNS name (Kubernetes Service or Docker network alias), not localhost. localhost inside the Bindery container is Bindery itself, blocked by the SSRF guard anyway.

---

Why no native Prowlarr type?

Prowlarr's value is aggregating indexer credentials and rewriting queries in a single place. Bindery consuming that aggregated feed via Newznab/Torznab is exactly the interface Prowlarr was designed to expose. A dedicated "Prowlarr" type in Bindery would add a second code path with no new functionality; it would also tie Bindery to Prowlarr's specific API quirks instead of the stable Newznab/Torznab spec that Jackett and other aggregators also speak.