Ableton Project Manager MCP

Tool · 2025

Model Context Protocol server for analyzing Ableton Live projects. Extracts missing samples, identifies plugins, detects duplicates by content, and returns structured JSON — accessible through any MCP-compatible AI assistant.

Features

  • Missing Sample Recovery: Extracts complete file paths from project XML. Paired with filesystem MCP, AI assistants can search drives and locate samples even if renamed or moved.
  • Plugin Inventory: Extracts complete plugin lists (VST3/VST2/AU) with instance counts. Distinguishes stock Ableton devices from third-party VSTs. Useful for system migrations and checking project portability.
  • BPM and Technical Metadata: Query projects by BPM, track count, arrangement length, automation lane count without opening Ableton. Structured JSON output suitable for scripting.
  • Duplicate Detection: Compares project content (track structure, device chains, MIDI patterns) rather than filenames. Identifies duplicates regardless of naming scheme.
  • Master Chain Analysis: Extracts and compares mastering device chains across projects. Groups projects by shared master processing.
  • Natural Language + Structured Data: Conversational queries return structured JSON. Example: "find my 140 BPM projects with Serum that have missing samples" returns JSON with file paths.

How It Works

The server provides seven MCP tools available to any MCP-compatible AI assistant:

  • scan_projects — Scans directory for .als files with basic info
  • analyze_projects — Deep analysis (BPM, tracks, plugins, samples, master chain)
  • find_recent — Recently modified projects
  • find_duplicates — Content-based duplicate detection
  • find_missing_plugins — Lists all plugins used across projects
  • analyze_master_chains — Groups projects by mastering setup
  • find_finished_projects — Classifies projects as finished vs sketches

Ableton .als files are gzipped XML. The server decompresses and parses them to extract metadata without requiring Ableton to be running. All output is structured JSON.

Safety First

Built with configurable safety constraints:

  • File limit: Stops after 100 .als files (configurable)
  • Size limit: Skips files over 50MB (configurable)
  • Depth limit: Only scans 3 folders deep (configurable)
  • Max analysis: 20 projects per call
  • Temp files always cleaned up, even on error

Use Cases

Example queries:

  • "Find all projects with missing samples and show me the file paths"
  • "Which projects use Serum? I'm migrating to a new computer"
  • "Show me all 140 BPM projects that only use stock plugins"
  • "Find duplicate projects - I have too many 'Final' versions"
  • "Which projects share the same mastering chain?"
  • "Export all project metadata to CSV for my spreadsheet"

Links