Overview
What is Six Hats MCP?
Six Hats MCP is a full implementation of Edward de Bono's Six Thinking Hats method using FastMCP and the OpenAI Agents SDK. It provides structured thinking tools for objective facts, emotions, risks, benefits, creativity, and orchestration.
How to use Six Hats MCP?
Set up an OpenAI API key in a .env file (or run python setup_env.py), then start the server with docker-compose up --build. Visit http://localhost:8000 to access the MCP endpoints. Six tools are available: white, red, black, yellow, green, and blue.
Key features of Six Hats MCP
- white: Objective facts & data
- red: Emotions & intuitions
- black: Risks & critiques
- yellow: Benefits & optimism
- green: Creativity & ideas
- blue: Manager & orchestration
Use cases of Six Hats MCP
—
FAQ from Six Hats MCP
Do I need an OpenAI API key to use Six Hats MCP?
Yes, the server requires an OpenAI API key. Provide it in a .env file as OPENAI_API_KEY=your_key_here.
How do I run Six Hats MCP locally?
Use docker-compose up --build to start the server. It will listen on port 8000.
What tools does Six Hats MCP expose?
It exposes six tools corresponding to de Bono's hats: white, red, black, yellow, green, and blue. Each hat has a specific thinking role.
How can I test the server?
Run poetry run pytest --cov=app to execute tests with coverage reporting.
What transport does Six Hats MCP use?
The server uses HTTP, with MCP endpoints available at http://localhost:8000.
