What is MCP? MCP (Model Context Protocol) lets you use this tutorial generator directly from Claude Desktop, making it as easy as asking "Generate a tutorial for this GitHub repo."

# Make sure you're in the main project directory
cd /path/to/Tutorial-Codebase-Knowledge
pip install -r requirements.txt

Copy the contents of mcp_config.json and add it to your Claude Desktop configuration:

On Mac: ~/Library/Application Support/Claude/claude_desktop_config.json On Windows: %APPDATA%/Claude/claude_desktop_config.json

Your config should look like this:

{
  "mcpServers": {
    "tutorial-codebase-knowledge": {
      "command": "python",
      "args": ["mcp_server_simple.py"],
      "cwd": "/path/to/Tutorial-Codebase-Knowledge/MCP",
      "env": {
        "PYTHONPATH": "/path/to/Tutorial-Codebase-Knowledge"
      }
    }
  }
}

Important: Update the "cwd" path to match your actual project location!

Add your preferred LLM API key to your environment:

# Choose one:
export GEMINI_API_KEY="your-key-here"
export OPENAI_API_KEY="your-key-here"  
export ANTHROPIC_API_KEY="your-key-here"
export OPENROUTER_API_KEY="your-key-here"

Restart Claude Desktop, then simply ask:

• "Generate a tutorial for https://github.com/user/repo"
• "Create documentation for my project at /path/to/local/project"
• "Make a tutorial for https://github.com/fastapi/fastapi"

The tutorial will be automatically saved to your docs/ folder with:

• index.md - Main overview with project summary
• 01_concept.md, 02_concept.md, etc. - Individual chapters
• Mermaid diagrams showing relationships

✅ Beginner-friendly tutorials explaining your codebase
✅ Multiple chapters breaking down key concepts
✅ Code examples with explanations
✅ Visual diagrams showing how components relate
✅ Saved to docs/ folder for easy access

"Tool not found": Make sure you restarted Claude Desktop after adding the config.

"Permission denied": Check that the cwd path in your config points to the correct MCP folder location.

"API error": Verify your API key is set correctly in your environment variables.

"No output": The tutorial generation takes 5-10 minutes. Be patient!

You can specify additional options in Claude:

• "Generate a tutorial in Spanish for [repo]"
• "Use OpenAI to generate a tutorial for [repo]"
• "Generate a tutorial and save it to ~/Desktop/my-docs"

• mcp_server_simple.py - Main MCP server (recommended)
• mcp_server.py - Full-featured MCP server
• mcp_config.json - Configuration template
• claude-desktop-config-snippet.json - Alternative config format