Skip to content
Engineering Decision Intelligence
How-toVersion 0.6.0

Install SwolfPy MCP

Create the qualified Python 3.10 environment and start the server locally or with Docker.

Local installation

bash
conda create -n swolfpy-mcp python=3.10 graphviz -y
conda activate swolfpy-mcp
git clone https://github.com/SDAI-institute/Swolfpy-mcp.git
cd Swolfpy-mcp/swolfpy-mcp
python -m pip install --upgrade pip
pip install -e ".[dev,runtime]"

Verify the package

bash
python -m pytest -q
python -m src

Docker

bash
docker compose up --build

The default compose configuration exposes the MCP endpoint at http://localhost:8031/mcp and health at http://localhost:8031/health.