lidarr-mcp

Python MCP (toolkit) to interact with a Lidarr instance. Designed to be used as a local "toolkit" for LLM integrations (e.g. Claude) or as a CLI. An optional FastAPI server is provided if you prefer HTTP tooling.

Goals

• Provide Python functions (tools) to list / get / create / update / delete artists, albums (releases) and tracks via Lidarr's API.

• Provide a CLI (Typer) to call these tools from the shell.

• Provide an optional FastAPI server to expose these tools over HTTP if needed.

Quick start

1. Clone the repo and create a virtualenv python -m venv .venv ..venv\Scripts\Activate.ps1 # or ..venv\Scripts\activate.bat

2. Install dependencies pip install -r requirements.txt

3. Copy .env.example to .env and set LIDARR_URL and LIDARR_API_KEY

4. Use the CLI python -m lidarr_mcp.cli artists list

How to plug this repo to Claude

• Option A (recommended local tooling): Run the CLI locally and configure Claude (or your LLM environment) so it can execute shell commands that call this CLI. For example, run the CLI via a controlled shell tool that your Claude instance can call.

• Option B (HTTP tool): Run the optional FastAPI server (python -m lidarr_mcp.server) and configure Claude to call the endpoints. This is provided for convenience but is not required.

Files created in this initial commit: CLI (Typer), lidarr client wrapper, tools module, optional FastAPI server, Dockerfile and docker-compose, .env.example.