Purpose
mcp-applemusic is a FastMCP server built for macOS, used to let AI agents control the local Apple Music. It is aimed at users who want to operate music playback, organize playlists, and view listening statistics through natural language, packaging various Apple Music operations into 74 tools that can be called by AI.
Main Capabilities
- Complete Apple Music control: Supports playback, volume adjustment, AirPlay level adjustment by device, EQ presets, repeat/shuffle modes, and MiniPlayer.
- Listening log and monthly Replay: Uses a local SQLite database and a background daemon scrobbler to continuously track listening records, and can generate Apple Replay reports.
- Cross-platform Spotify importer: Can directly import public Spotify playlists and albums into Apple Music, with no login or token required.
- Smart AI DJ transitions: Automatically calculates BPM and trims start/end offsets to achieve continuous seamless crossfades.
- Hybrid lyrics and universal links: Instantly retrieves lyrics through LRCLIB, and generates song.link universal links, making it easy to share songs between Spotify/YouTube/Tidal.
- Playlist management and export: Supports nested folders, deduplication, automatic sorting, and can export to JSON, CSV, or Markdown.
How to Use
Runtime requirements:
- macOS with the Apple Music app installed.
- uv package manager (can be installed via
brew install uv).
Configuration methods include:
- Method 1: One-click automatic installation via Smithery (for automatic installation and configuration of Claude Desktop).
- Method 2: Claude Desktop: Add the configuration in
~/Library/Application Support/Claude/claude_desktop_config.json. - Method 3: Cursor / Windsurf / VS Code: Add the configuration in the MCP settings (
.cursor/mcp.jsonor Cline/Roo settings); it can also be configured via Git URL, for example"args": ["git+https://github.com/suleyman416/mcp-applemusic.git"].
Permission note: On first run, macOS will request permission for the AI client or terminal to control Music.app. Please click Allow, or go to System Settings ➔ Privacy & Security ➔ Automation to confirm.
After connecting, you can give instructions to the AI agent in natural language, for example:
- Playback and audio: “Play Location by Dave” “Set volume to 80% and switch EQ to Hip-Hop” “Switch audio output to my AirPods Pro” “Turn on shuffle for my Gym playlist”
- Discovery and conversion: “Import this Spotify playlist: https://open.spotify.com/playlist/... "Generate a universal share link for what's currently playing" "Show me the UK Top 10 Apple Music chart" "Get the lyrics for this song"
- Analysis and Replay: "Give me my August Replay report" "What is my music listening personality profile?" "Show my listening history for the last 10 songs"
- DJ and library engineering: "Configure smart DJ crossfades for my 'Party' playlist" "Find and remove duplicate songs in my 'Favorites' playlist" "Export my 'Chill Vibes' playlist to Markdown"
Use cases
- Users who want to control Apple Music playback and volume on their Mac through a conversational AI.
- Users who need to migrate Spotify playlists to Apple Music.
- Users interested in listening statistics, monthly Replay reports, and listening history analysis.
- Users who need automatic DJ crossfades and playlist deduplication, sorting, and export in scenarios such as parties.
Limitations
- Only macOS is supported, and the Apple Music app must already be installed.
- Requires the uv package manager.
- On first run, system permission to automate control of Music.app must be granted.
- Spotify import only supports public playlists and albums, and requires no login or token.
- Data runs and is stored 100% locally via stdio, and no credentials, tokens, or personal identifiers are stored or transmitted.
- Licensed under MIT.
Other details (such as the full list of 74 tools, development and contribution steps, and the full text of PRIVACY.md) are not yet provided.
