synthlab
AI-powered synthesis lab — Compose songs, generate Pure Data & VCV Rack patches, map MIDI controllers, and control live synths through Claude
claude mcp add --transport stdio j0kz-synthlab-mcp-server npx -y synthlab-mcp-server
How to use
synthlab-mcp-server is an MCP (Model Context Protocol) server that turns Claude into a full synthesis workstation. It provides 10 tools plus a prompt to generate and manipulate Pure Data (Pd) patches, create VCV Rack patches, analyze and validate audio graphs, and map MIDI controllers to rack parameters. The server also includes a Pd Core Engine for parsing and serializing Pd patches, a template engine with 11 preset instruments, and a rack builder for wiring modules and MIDI controllers. You can request generated Pd or VCV Rack content by describing the desired instrument, patch, or rack configuration, and the server will return the corresponding patch files, templates, or wiring instructions ready to open in their respective hosts. Use cases include composing songs from genre descriptions, creating multi-module Pd racks, auto-wiring MIDI controllers like Korg nanoKONTROL2, Arturia MicroFreak, and Roland TR-8S, and exporting VCV Rack patches compatible with VCV Rack 2.x.
To use it, run the MCP server with the provided command, then interact with Claude (or any MCP client) to issue commands like: create a Pd rack with clock and sequencer, generate a VCV Rack patch with a VCO → VCF → VCA chain, or map a hardware controller to a Pd module parameter. The tools will produce outputs such as .pd files, .vcv patches, and wiring instructions that you can load directly into the corresponding environments. The server emphasizes zero runtime dependencies beyond the MCP SDK and Zod, ensuring a lightweight and extensible workflow for synthesis, patch design, and live controller integration.
How to install
Prerequisites:
- Node.js (LTS version, e.g., 18.x or 20.x) and npm.
- Basic familiarity with MCP workflow and the host applications (Pure Data and VCV Rack).
Option A: Run directly with npx (no local install required)
- Ensure Node.js and npm are installed:
- node -v
- npm -v
- Start the MCP server via npx (this fetches the package temporarily): npm install -g npm@latest npx -y synthlab-mcp-server
Option B: Install locally via npm (preferred for development)
- Clone the repository or ensure you have access to the npm package
- Install locally: npm install synthlab-mcp-server --save
- Run the server (adjust path if the package exposes a CLI): npx synthlab-mcp-server
Option C: Docker (if you prefer containerized setup)
- Pull the image (if provided by the project): docker pull synthlab/synthlab-mcp-server:latest
- Run the container: docker run -i synthlab/synthlab-mcp-server:latest
Notes:
- If your environment restricts network access, you may need to configure npm/yarn proxy settings or use a local registry.
- If using a non-default port or environment variables, refer to the project’s environment configuration options to customize runtime behavior.
Additional notes
Environment variables and configuration tips:
- MCP server may expose environment variables for API keys or model endpoints used by Claude integration; set these as VAR_NAME=value in your runtime environment.
- If you encounter validation or patch generation errors, verify that your Pd/Vivado paths are correctly configured and that the JSON spec for patches/racks adheres to the server’s expected schema.
- For MIDI controller mappings, ensure the host system has the correct permissions to access USB devices and that the controller mappings align with the available modules in the rack templates.
- When upgrading, re-run any required schema migrations if the MCP SDK version changes, and consult the changelog for breaking changes to tool interfaces.
Related MCP Servers
ls
List MCP Server configurations in your system used by AI applications like Cursor, Claude Desktop, VS Code and others
CanvasMCPClient
Canvas MCP Client is an open-source, self-hostable dashboard application built around an infinite, zoomable, and pannable canvas. It provides a unified interface for interacting with multiple MCP (Model Context Protocol) servers through a flexible, widget-based system.
mcp -amazon
🛍📦 Unofficial Amazon Model Context Protocol Server (MCP) - Search products and purchase directly from Claude AI! ✨
mcp-web-search-tool
A MCP server providing real-time web search capabilities to any AI model.
vscode-context
MCP Server to Connect with VS Code IDE
obsidian-ai-curator
Smart note consolidation for Obsidian using Claude AI. Automatically identifies and merges scattered fragments into structured knowledge via MCP. Like Tetris for your thoughts - pieces fall into place.