Sonic Pi MCP
Interact with Sonic Pi, the live coding music synth, using OSC messages.
Sonic Pi MCP
A Model Context Protocol (MCP) server that allows AI assistants to interact with Sonic Pi through OSC messages. This enables AI tools like Claude and Cursor to create music and control Sonic Pi programmatically.
Features
- Play individual notes with customizable synth parameters
- Execute arbitrary Sonic Pi code
- Works with any MCP-compatible client (Claude Desktop, Cursor, etc.)
Prerequisites
Sonic Pi Configuration
Before using the MCP server, you need to add the following code to your Sonic Pi buffer. This code handles the OSC messages sent by the server:
# Required Sonic Pi configuration
# Add this to a buffer in Sonic Pi and run it
live_loop :code_runner do
use_real_time
code = sync "/osc*/run-code"
# Since we receive the code as a string, we can use eval to execute it
# The code comes as the first element of the message
begin
eval(code[0].to_s)
rescue Exception => e
puts "Error executing code: #{e.message}"
end
end
Make sure this code is running in Sonic Pi before using the MCP server.
You can also try running this script in SonicPi to have AI assitants play songs in a queue.
Integration with Clients
Cursor
Add to ~/.cursor/mcpServers.json:
{
"mcpServers": {
"sonic_pi_mcp": {
"name": "Sonic Pi MCP",
"command": "bunx",
"args": ["sonic-pi-mcp"],
"transport": {
"type": "stdio"
}
}
}
}
Claude Desktop
Add to ~/Library/Application Support/Claude/claude_desktop_config.json:
{
"mcpServers": {
"sonic_pi_mcp": {
"command": "bunx",
"args": ["sonic-pi-mcp"],
}
}
}
Available Tools
play_note
Plays a single note with customizable parameters.
Parameters:
note(required): MIDI note number (0-127)synth(optional): Synth to use (e.g., ":saw", ":beep", ":prophet")sustain(optional): Note duration in seconds (default: 1)cutoff(optional): Filter cutoff frequency (default: 100)
Example:
// Play middle C with saw wave synth
{
"name": "play_note",
"parameters": {
"note": 60,
"synth": ":saw",
"sustain": 0.5,
"cutoff": 80
}
}
run_code
Executes arbitrary Sonic Pi code.
Parameters:
code(required): Sonic Pi code to execute
Example:
{
"name": "run_code",
"parameters": {
"code": "use_synth :prophet\nplay_pattern_timed [60, 64, 67], [0.5]"
}
}
Example Usage
Here are some example interactions using the MCP tools:
Simple Melody
// Play a C major arpeggio
{
"code": `
use_synth :piano
play_pattern_timed [60, 64, 67, 72], [0.25], release: 0.1
`
}
Complex Pattern
// Create a rhythmic pattern
{
"code": `
live_loop :rhythm do
use_synth :tb303
play choose(chord(:C3, :minor)), release: 0.2, cutoff: rrand(60, 120)
sleep 0.25
end
`
}
Troubleshooting
-
No Sound
- Ensure Sonic Pi is running
- Check that the OSC handler code is running in Sonic Pi
- Verify Sonic Pi is listening on port 4560 (default)
-
Connection Errors
- Check if another instance of the server is running
- Restart Sonic Pi
- Ensure no other applications are using port 4560
-
Code Execution Errors
- Check the Sonic Pi log window for error messages
- Verify the syntax of your Sonic Pi code
- Ensure all required synths and samples are available
Development
# Clone the repository
git clone https://github.com/abhishekjairath/sonic-pi-mcp.git
cd sonic-pi-mcp
# Install Bun if you haven't already
curl -fsSL https://bun.sh/install | bash
# Install dependencies
bun install
# Start Sonic Pi and run the OSC handler code (see Sonic Pi Configuration section)
# Start the server in development mode
bun run dev
Testing with MCP Inspector
- Install and start the MCP Inspector:
npm install -g @modelcontextprotocol/inspector
mcp-inspector
-
Open your browser and navigate to http://localhost:3000
-
In the MCP Inspector UI, configure the connection:
- Command:
bun - Arguments:
run src/server.ts - Working Directory:
/path/to/your/sonic-pi-mcp(use your actual project path) - Transport Type: stdio
- Command:
-
Test the
play_notetool:
{
"name": "play_note",
"parameters": {
"note": 60,
"synth": ":beep",
"sustain": 0.5
}
}
- Test the
run_codetool:
{
"name": "run_code",
"parameters": {
"code": "use_synth :prophet\nplay_pattern_timed scale(:c4, :major), [0.25]"
}
}
Troubleshooting Development Issues
-
Bun Installation Issues
- Make sure Bun is in your PATH
- Try running
bun --versionto verify the installation - If using Claude Desktop, use the full path to Bun in the config
-
MCP Inspector Connection Issues
- Verify the server is running (
bun run dev) - Check that the working directory path is correct
- Ensure no other instances of the server are running
- Verify the server is running (
-
OSC Communication Issues
- Confirm Sonic Pi is running and the OSC handler code is active
- Check the server logs for connection errors
- Verify port 4560 is available and not blocked
Contributing
- Fork the repository
- Create your feature branch (
git checkout -b feature/amazing-feature) - Commit your changes (
git commit -m 'Add some amazing feature') - Push to the branch (
git push origin feature/amazing-feature) - Open a Pull Request
License
This project is licensed under the MIT License - see the LICENSE file for details.
Похожие серверы
Scout Monitoring MCP
спонсорPut performance and error data directly in the hands of your AI assistant.
Alpha Vantage MCP Server
спонсорAccess financial market data: realtime & historical stock, ETF, options, forex, crypto, commodities, fundamentals, technical indicators, & more
Raysurfer Code Caching
MCP server for LLM output caching and reuse. Caches and retrieves code from prior AI agent executions, delivering cached outputs up to 30x faster.
consult7
Analyze large codebases and document collections using high-context models via OpenRouter, OpenAI, or Google AI -- very useful, e.g., with Claude Code
GDB MCP Server
An MCP server that enables LLM clients to interact with GDB for debugging and binary analysis.
Remote MCP Server Authless
An example of a remote MCP server deployable on Cloudflare Workers without authentication.
Zeek-MCP
Integrates Zeek network analysis with conversational AI clients. Requires an external Zeek installation.
Squidler.io
Squidler is designed to validate your web app as a human based on natural language use cases, without write brittle, DOM-dependent tests.
Vibes
Transforms Claude Desktop into a conversational development environment using distributed MCP servers.
DeepSeek MCP Server
An MCP server for the DeepSeek API, providing code review, file management, and account management.
Elementor MCP Server
Perform CRUD operations on Elementor page data for a target WordPress website.
EdgeOne Pages MCP
An MCP server and client implementation for EdgeOne Pages Functions, supporting OpenAI-formatted requests.