carla

The Carla MCP Server provides a production-ready interface to control the Carla audio plugin host via natural language, exposing 45 tools across 7 categories to manage sessions, plugins, routing, automation, analysis, and hardware integration. It enables an AI assistant (LLM) to perform complex audio production tasks—from loading projects and configuring plugins to real-time analysis and mix optimization—through structured tool calls and natural language prompts. Use the included MCP client configuration to connect an LLM-enabled client and begin issuing high-level requests like loading a project, configuring a drum bus with compression, or setting up a live-performance session with low latency. The server integrates MixAssist resources for professional mixing guidance and provides real-time monitoring and route visibility to keep workflows smooth in demanding environments.

Prerequisites:

• Carla Audio Plugin Host installed (Carla production-ready build preferred).
• Python 3.12 or newer.
• JACK Audio Connection Kit (recommended) or compatible audio backend.

Installation steps:

1. Install Carla and Python dependencies:

# Install Carla (example for Debian/Ubuntu)
sudo apt update
sudo apt install carla carla-dev

# Install Python and virtual environment tools if needed
sudo apt install python3.12 python3-venv python3-pip

2. Clone the MCP server repository and install Python dependencies:

git clone https://github.com/your-org/carla-mcp-server.git
cd carla-mcp-server

# Install project dependencies
python3 -m pip install -r requirements.txt

# Optional development install
python3 -m pip install -e ".[dev]"

3. Configure environment variables (adjust paths as appropriate):

export CARLA_PATH=/usr/share/carla
export PYTHONPATH=$CARLA_PATH/source/frontend:$PYTHONPATH
export LD_LIBRARY_PATH=/usr/lib/carla:$LD_LIBRARY_PATH

4. Run basic tests and verify Carla integration:

pytest tests/
python -c "import carla_backend; print('Carla integration working!')"

5. Start the MCP server via the standard command:

python server.py
# or with debugging
python server.py --debug --log-level INFO

Optional MixAssist setup (recommended):

python setup_mixassist.py --download

This downloads the MixAssist dataset, creates the .env configuration, and enables MixAssist MCP resources. For full details, refer to MIXASSIST_SETUP.md.

Environment variables and paths may vary by system. Ensure Carla paths, JACK settings, and Python paths align with your installation. If you encounter import or plugin loading errors, verify that CARLA_PATH points to a valid Carla installation and that LD_LIBRARY_PATH includes Carla libraries. The MCP server is designed for production workflows; enable verbose logging during setup to capture actionable error details. When connecting an MCP client, ensure the client uses the carla-mcp server name and that the environment variables are exposed to the client process if needed. Consider running the server inside a virtual environment to isolate dependencies and simplify upgrades.