arr-mcp¶
Arr Suite API + MCP Server + A2A Agent for the agent-utilities ecosystem — a typed, deterministic control surface over Sonarr, Radarr, Lidarr, Prowlarr, Bazarr, Seerr, and Chaptarr.
Official documentation
This site is the canonical reference for arr-mcp, maintained alongside every
release.
Overview¶
arr-mcp wraps the REST surface of the Arr Suite of media-automation services
with consolidated, action-routed MCP tools and a built-in Pydantic-AI agent. It
provides:
- Per-service API clients — request-session facades over Sonarr, Radarr, Lidarr, Prowlarr, Bazarr, Seerr, and Chaptarr, each constructed from the environment.
- Action-dispatch MCP tools — one consolidated tool per service
(
sonarr_action,radarr_action, …) that keeps the LLM tool surface compact and togglable. - An A2A agent server (
arr-agent) that calls the MCP tools over the Agent Control Protocol with an optional web interface.
Each service connector remains inactive when its credentials are absent, so the server runs cleanly with only the integrations you configure.
Explore the documentation¶
- Installation — pip, source, extras, and the prebuilt Docker image.
- Deployment — run the MCP and agent servers, Docker Compose, Caddy + Technitium.
- Usage — the MCP tools, the Python API clients, and the CLI.
- Backing Platform — provision the Arr Suite services with Docker.
- Overview — ecosystem role, enterprise readiness, and configuration.
- Concepts — the
CONCEPT:ARR-*registry.
Quick start¶
Connect it to your Arr Suite services:
export SONARR_BASE_URL=http://localhost:8989
export SONARR_TOKEN=your_sonarr_api_key
arr-mcp --transport streamable-http --host 0.0.0.0 --port 8000
See Installation and Deployment for the full matrix (PyPI extras, Docker image, all transports, the agent server, reverse proxy, DNS).