User Setup Guide (Clean)
June 29, 2026 · View on GitHub
Compatibility — Requires Lidarr 3.0.0.4855+ on the plugins/nightly branch. See the README compatibility notice before enabling Brainarr.
How to use this guide
- Finish the README quick start so Brainarr is built, installed, and visible inside Lidarr.
- Consult the Brainarr AI Provider Guide and wiki provider pages to pick your primary and fallback models.
- Use the steps below inside Lidarr to wire everything together and validate the installation.
Step 1 — Prepare the environment
- Run
./setup.ps1(Windows) or./setup.sh(macOS/Linux) from the repository root to fetch Lidarr assemblies and restore the Brainarr solution. - Optional:
pwsh ./build.ps1 --testto run the full validation suite before deploying to production. - Keep
ext/Lidarr/_output/net8.0intact—this is where the setup scripts place the assemblies Lidarr expects.
Step 2 — Choose and configure a provider
- Reference the Provider status matrix for current verification notes.
- Follow the matching wiki article for authentication details and rate-limit guidance: Local Providers or Cloud Providers.
- Record API keys using the secure storage that matches your operating system (Keychain, Credential Manager, libsecret, etc.).
Step 3 — Add the Brainarr import list in Lidarr
- In Lidarr, go to Settings → Import Lists → Add (+).
- Select Brainarr and choose a descriptive name (e.g.,
AI Music Recommendations). - Set Enable Automatic Add to Yes unless you plan to gate every recommendation manually.
- Pick the Quality Profile, Metadata Profile, and Root Folder that match your library.
- Add any tags (for example
ai-recommendations) so you can filter lists later. - Save to create the list—the plugin will render provider-specific settings underneath.
Step 4 — Wire up providers and test connectivity
- In the provider panel, choose your Primary Provider and optional Fallback Providers.
- Supply required credentials or base URLs. Use local endpoints (
http://localhost:11434,http://localhost:1234, etc.) for Ollama/LM Studio. - Click Test. A green toast confirms Brainarr can authenticate and query the provider. Resolve failures using the troubleshooting links in the UI or the wiki.
- (Optional) Enable additional providers and set their priorities if you want automatic failover—Brainarr will advance through enabled providers in ascending priority when the primary errors or exceeds quotas.
Step 5 — Request the first recommendations
- Open Import Lists → Brainarr → Manual and trigger Fetch.
- Inspect the generated recommendations; approve the albums you want to monitor.
- Schedule automatic refreshes in Import Lists → Options → Interval once you are satisfied with the output.
- Track provider usage and headroom via Lidarr System → Logs (
Brainarr:entries) and your provider dashboards.
Ongoing operations
- Review queue workflow: Follow the “Operations” section in the wiki to triage recommendations, invalidate cache entries, and monitor prompt metrics.
- Observability: Metrics and log field definitions are centralised in the wiki’s Observability & Metrics page.
- Upgrades: When updating Brainarr, rerun the setup script and review the CHANGELOG for migrations or new settings.
Troubleshooting
- Start with
docs/troubleshooting.mdfor common failure modes (authentication, token budgets, cache states). - Provider-specific issues (429/401 responses, model discovery failures) are linked from the wiki provider pages.
- If Brainarr fails to load, confirm your Lidarr branch matches the README requirement and check System → Logs for
Brainarr: minVersionmessages.
Keep this guide focused on the user-facing workflow. Any time a step changes (new settings, provider parameter, etc.), update docs/providers.yaml or the relevant wiki page so every surface stays aligned.