CDP MCP Server
Access Composers' Desktop Project (CDP) sound transformation programs. Requires a separate CDP installation.
CDP MCP Server
A Model Context Protocol (MCP) server that provides direct access to the Composers' Desktop Project (CDP) sound transformation programs. This server offers an ultra-rigid workflow with zero interpretation, exposing CDP's raw functionality through simple, reliable tools.
Overview
CDP MCP Server v7 implements a minimalist approach to CDP integration:
- Direct execution - No command parsing or interpretation
- Raw usage text - See exactly what CDP shows
- Simple tools - Just 6 core functions
- Data file support - Create parameter files for complex operations
Features
- 🎵 Full CDP Access - Execute any CDP program with complete control
- 📚 Program Discovery - List all available CDP programs by category
- 📖 Usage Information - Get raw usage text directly from CDP
- 📄 Data File Creation - Create parameter files required by CDP programs
- 🎛️ Spectral Preparation - Helper for PVOC analysis
- 📊 Sound Analysis - Get basic properties of audio files
Requirements
System Requirements
- macOS, Linux, or Windows
- Python 3.8 or higher
- CDP (Composers' Desktop Project) installed
Python Dependencies
mcp
soundfile
numpy
CDP Installation
- Download CDP from the official website
- Install CDP following the platform-specific instructions
- Set the
CDP_PATHenvironment variable to your CDP programs directory:export CDP_PATH="/path/to/cdp/programs"
Installation
1. Clone the Repository
git clone https://github.com/DavidPiazza/CDP_MCP.git
cd CDP_MCP
2. Install Dependencies
pip install -r requirements.txt
Or install individually:
pip install mcp soundfile numpy
3. Configure CDP Path
Set your CDP installation path:
export CDP_PATH="/Users/yourname/cdpr8/_cdp/_cdprogs" # macOS example
4. Run the Server
python CDP_MCP_v7.py
MCP Client Configuration
To use this server with an MCP client (like Claude Desktop), add to your configuration:
{
"mcpServers": {
"cdp": {
"command": "python",
"args": ["/path/to/CDP_MCP_v7.py"],
"env": {
"CDP_PATH": "/path/to/cdp/programs"
}
}
}
}
Usage
Basic Workflow
-
List Available Programs
list_cdp_programs() # Returns categorized list of all CDP programs -
Get Program Usage
get_cdp_usage('blur') # Returns exact CDP usage text -
Execute Commands
execute_cdp(['blur', 'blur', 'input.ana', 'output.ana', '50']) # Direct execution with no interpretation
Example Operations
Time Stretch
# Check usage
get_cdp_usage('stretch')
# Execute
execute_cdp(['stretch', 'time', '1', 'input.ana', 'output.ana', '2.0'])
Spectral Blur
# Note the double syntax for blur
execute_cdp(['blur', 'blur', 'input.ana', 'output.ana', '50'])
Granular Synthesis
execute_cdp(['modify', 'brassage', '4', 'input.wav', 'output.wav', '0.02', '-0.5', '-r200'])
Using Data Files
# Create data file for tesselate
create_data_file('tess_data.txt', '5 5 5 5\n0.0 0.1 0.2 0.3')
# Execute with data file
execute_cdp(['tesselate', 'tesselate', '1', 'in1.wav', 'in2.wav', 'in3.wav', 'in4.wav', 'output.wav', '0.5', 'tess_data.txt'])
Tools Reference
list_cdp_programs()
Lists all available CDP programs organized by category (Spectral Processing, Time Domain, Synthesis, etc.)
get_cdp_usage(program, subprogram=None)
Returns raw usage information for a CDP program by running it without arguments.
execute_cdp(command)
Executes a CDP command given as an array of strings. No parsing or interpretation.
create_data_file(filepath, content)
Creates text data files required by certain CDP programs.
prepare_spectral(input_file, output_file, window_size=2048)
Helper function to perform PVOC analysis for spectral processing.
analyze_sound(filepath)
Returns basic properties of a sound file (duration, sample rate, channels, etc.)
Architecture
The server follows an ultra-rigid design philosophy:
- No command parsing - Commands are arrays, not strings
- No parameter validation - CDP handles all validation
- No output interpretation - Raw CDP output is returned
- Minimal abstraction - Direct CDP access only
Tips
- Always check usage with
get_cdp_usage()before executing - CDP may return non-zero exit codes even on success
- Check if output files exist to verify successful execution
- Use exact command arrays - the server does no interpretation
- Create data files when CDP usage mentions DATAFILE requirements
Troubleshooting
CDP Not Found
Ensure CDP_PATH environment variable points to your CDP programs directory.
Apple Silicon Issues
The server automatically handles x86_64 emulation on Apple Silicon Macs using arch -x86_64.
Command Failures
- Check the exact usage with
get_cdp_usage() - Verify all file paths exist
- Ensure proper command array format
- Check CDP's stderr output for specific errors
License
MIT License - See LICENSE file for details
Acknowledgments
- Composers' Desktop Project for the amazing sound transformation tools
- Model Context Protocol for the MCP framework
Related Servers
Pica MCP Server
An MCP server for Pica that enables seamless interaction with various third-party services through a standardized interface.
x64dbgMCP
An MCP server that connects LLMs with the x64dbg debugger, enabling natural language control over debugging functions.
MCP-guide
A guide for setting up an MCP server using a Python virtual environment and integrating it with the Cline VS Code extension.
MCP Proxy Server
A proxy server for aggregating and serving multiple MCP resource servers through a single endpoint.
Distance Tools MCP
A remote MCP server example deployable on Cloudflare Workers, featuring customizable tools and no authentication.
Azure DevOps MCP Server for Cursor
An MCP server for Azure DevOps with tools for project management, work items, pull requests, builds, tests, and more.
Claude Code Exporter
Export and organize Claude Code conversations with powerful filtering. Supports CLI and MCP server integration for Claude Desktop.
UnrealMCP Plugin
An unofficial MCP server plugin for remote control of Unreal Engine using AI tools.
Tiktoken MCP
Count tokens using OpenAI's tiktoken library.
Storyblok MCP Server
Manage your Storyblok CMS using natural language through AI tools.