Reaper MCP Server
This is an MCP server that connects Reaper projects to an MCP client like Claude Desktop, enabling you to ask questions about your projects and get comprehensive audio analysis for mixing feedback.
Tools
Project Discovery & Parsing
find_reaper_projects: Finds all Reaper projects in the directory you specified in the config.parse_reaper_project: Parses a Reaper project file (.RPP) and returns detailed information including tempo, tracks, FX chains, and audio items.
These tools work in tandem. When you ask Claude a question about a specific Reaper project, it will use the find_reaper_projects tool to find the project, then use the parse_reaper_project tool to parse the project and answer your question.
Installed FX Discovery
-
list_installed_fx(plugin_type=None, search_query=None): Lists all installed FX/plugins available in Reaper.Parameters:
plugin_type(optional): Filter by plugin type (VST2, VST3, AU, JS, CLAP)search_query(optional): Search plugins by name, manufacturer, or type
Returns: List of installed plugins including:
- Plugin name
- Plugin type (VST2, VST3, AU, JS, CLAP)
- File path
- Manufacturer (when available)
Example Questions:
- "What synth plugins do I have installed?"
- "Show me all my Waves plugins"
- "I'm looking for a warbly synth. What options do I have from my already installed plugins?"
- "List all my VST3 plugins"
- "Do I have any reverb plugins?"
- "What iZotope plugins do I have?"
- "Show me all my Audio Unit plugins"
Note: This tool scans your Reaper plugin cache files. If you recently installed new plugins and haven't scanned them in Reaper yet, they won't appear in the results. Make sure to open Reaper and let it scan for new plugins first.
Audio Analysis
-
analyze_audio_files(project_path, track_filter=None): Analyzes all audio files in a Reaper project for mixing feedback.Parameters:
project_path(required): Path to the .RPP project filetrack_filter(optional): Filter tracks by name (e.g., "Vocal" to analyze only vocal tracks)
Returns: Comprehensive audio analysis including:
- Level Analysis: Peak levels, RMS, clipping detection
- Frequency Analysis: Spectral content, energy distribution across frequency bands
- Stereo Imaging: Stereo width, phase coherence, mono compatibility
- Dynamic Range & Loudness: LUFS (loudness standards), true peak, crest factor
Example Questions:
- "Analyze all audio in my Rock Song project"
- "Check the vocal tracks for clipping"
- "Is my mix too loud for streaming platforms?"
- "Are there any phase issues in my drum tracks?"
Warning Thresholds:
- Peak > -0.3 dBFS: Risk of clipping
- Clipping detected: Digital distortion present
- Excessive low frequency energy (> -6 dB): Muddy mix
- Phase coherence < 0.5: Phase cancellation issues
- LUFS > -8: Too loud for streaming (Spotify target: -14 LUFS)
- Crest factor < 6 dB: Possibly over-compressed
To see all data structures parsed from projects, check out the src/reaper_mcp_server/reaper_dataclasses.py file.
Setup
-
Install Dependencies
uv venv source .venv/bin/activate uv pip install . -
Configure Claude Desktop
- Follow the instructions to configure Claude Desktop for use with a custom MCP server
- Find the sample config in
setup/claude_desktop_config.json - Update the following paths in the config:
- Your
uvinstallation path - Your Reaper project directory
- This server's directory
- Your
-
Launch and Configure
- Open Claude Desktop
- Click the '+' icon on the chat box
- Click on 'Connectors' and you should see the 'reaper' connector enabled

-
Ask Away!
- Ask questions about your Reaper project
- Always include the name of the specific Reaper project you're asking about
- You can expand the tool boxes to see the raw project data being passed to Claude

Recommend MCP Servers 💡
devhub-cms-mcp
An MCP server that integrates DevHub CMS with Large Language Models, enabling LLMs to manage content, businesses, locations, and media within the DevHub CMS system.
danielscholl/backlog-manager-mcp
A simple task tracking and backlog management MCP server for AI assistants supporting SSE and stdio transports
ipfs-uploader
A TypeScript-based MCP server for uploading images to IPFS and managing a simple notes system, demonstrating core MCP concepts with resources, tools, and prompts.
ai-coder-tools
Provides a collection of Model Context Protocol (MCP) server implementations, including tools for Azure DevOps, Git, Kusto, and general search, designed to extend Roo Code capabilities for AI-assisted coding.
@larryhudson/linear-mcp-server
An MCP server that acts as a bridge between Claude and the Linear task/issue management system
things3-mcp
An MCP server that provides comprehensive integration with Things3 on macOS, enabling AI assistants to manage tasks, projects, and tags with intelligent error correction and automatic tag creation.