Claude Auto Documenter V2

MCP Server Node.js Version Python Version License: MIT

An MCP (Model Context Protocol) server that provides automatic documentation generation for Claude Code. Integrates with your development workflow to capture and document project implementations.

🌟 Features

🎯 Core Capabilities

• 🔮 Automatic Documentation Generation: Captures project implementations automatically
• 🎪 Simulation Capabilities: Test documentation triggers with simulation tools
• 🌊 File System Watching: Real-time documentation updates on file changes
• 🌍 Multi-Project Support: Manages documentation for multiple projects simultaneously
• 🎨 Intelligent Context Extraction: Automatically detects project type and technologies
• ⚖️ Queue Management: Robust queuing system for documentation tasks

🚀 Integrations

• ⚡ Claude Code Integration: Native MCP server for seamless Claude Code workflows
• 🔬 Technology Detection: Supports Node.js, Python, Rust, Go, and general projects
• 💎 Configurable Storage: Flexible documentation storage with environment variables
• 🏆 Open Source: MIT licensed for community use and collaboration

📋 Prerequisites

• 🟢 Node.js: ≥18.0.0
• 🐍 Python: ≥3.8 (for monitoring scripts)
• 🎯 Claude Code: Latest version with MCP support

🛠️ Installation

🚀 Option 1: Quick Install (Recommended)

# Clone the repository
git clone https://github.com/Puneet8800/claude-auto-documenter-v2.git
cd claude-auto-documenter-v2

# Run the installation script
chmod +x scripts/install.sh
./scripts/install.sh

🔧 Option 2: Manual Installation

# Install Node.js dependencies
npm install

# Install Python dependencies
pip install -r requirements.txt

# Make scripts executable
chmod +x scripts/*.sh

⚙️ Configuration

🌍 Environment Variables

Create a .env file in the project root:

# Documentation storage path (default: ~/auto-docs)
AUTO_DOC_DOCS_PATH=/path/to/your/documentation

# Additional configuration options (optional)
# AUTO_DOC_MAX_RETRIES=3
# AUTO_DOC_RETRY_DELAY=1000

# Claude agents directory (default: ~/.claude/agents)
CLAUDE_AGENTS_PATH=/path/to/claude/agents

# Retry configuration
AUTO_DOC_MAX_RETRIES=3
AUTO_DOC_RETRY_DELAY=1000

# File watching patterns (comma-separated)
AUTO_DOC_WATCH_PATTERNS="**/*.js,**/*.ts,**/*.py,**/*.md,**/*.json"
AUTO_DOC_IGNORE_PATTERNS="node_modules/**,.git/**,dist/**,build/**"

🎯 Claude Code Integration

Add to your Claude Code MCP configuration (claude_desktop_config.json):

{
  "mcpServers": {
    "auto-documenter": {
      "command": "node",
      "args": ["/path/to/auto-documenter/src/index.js"],
      "env": {
        "AUTO_DOC_DOCS_PATH": "/your/documentation/path"
      }
    }
  }
}

🎯 Usage

🧰 MCP Tools Available

Once configured, these tools are available in Claude Code:

🎬 trigger-auto-documentation

Manually trigger documentation generation for a project.

// Example usage in Claude Code
{
  "projectName": "my-project",
  "action": "document", // or "update", "init", "agent-completion", "command-completion"
  "workingDir": "/path/to/project"
}

📈 get-documentation-status

Get current status of the documentation system.

{
  "projectName": "my-project" // optional - get status for specific project
}

👀 setup-file-watching

Enable or disable file system watching for automatic documentation.

{
  "workingDir": "/path/to/project",
  "enable": true
}

🎭 simulate-agent-completion

Test automatic documentation with simulated agent completion.

{
  "agentName": "test-agent",
  "taskDescription": "Implemented new feature",
  "workingDir": "/path/to/project"
}

⚡ simulate-command-completion

Test automatic documentation with simulated command completion.

{
  "command": "npm test",
  "output": "All tests passed",
  "workingDir": "/path/to/project"
}

🎣 Documentation Triggers

The system can generate documentation through:

1. 📂 File Changes: When monitored files are modified (if file watching is enabled)
2. 🎤 Manual Triggers: Using MCP tools to generate documentation
3. 🧪 Simulation Tools: Test agent and command completion scenarios

💻 Using the MCP Tools

Once configured in Claude Code, you can use the available MCP tools:

• trigger-auto-documentation: Manually trigger documentation generation
• get-documentation-status: Check the current status
• setup-file-watching: Enable/disable file monitoring
• simulate-agent-completion: Test agent completion scenarios
• simulate-command-completion: Test command completion scenarios

🏗️ Architecture

graph TD
    A[Claude Code] --> B[Auto-Documenter MCP Server]
    B --> C[Task Completion Detector]
    B --> D[File System Watcher]
    B --> E[Documentation Pipeline]
    
    C --> F[Queue Manager]
    D --> F
    E --> F
    
    F --> G[Project Context Extractor]
    G --> H[Documentation Generator]
    H --> I[Storage Manager]
    
    I --> J[Project Documentation Files]
    
    subgraph "Triggers"
        K[Agent Completion]
        L[Command Execution]
        M[File Changes]
    end
    
    K --> C
    L --> C
    M --> D

🔧 Key Components

• 🎯 MCP Server (src/index.js): Main server handling Claude Code integration
• 🕵️ Task Completion Detector (src/task_completion_detector.py): Monitors for task completion events
• 📝 Documentation Pipeline (src/documentation_pipeline.py): Processes and generates documentation
• 👁️ File System Watcher: Real-time monitoring of project files
• ⚖️ Queue Manager: Robust queuing system for documentation tasks

🧪 Testing

🏃‍♂️ Run Tests

# Run JavaScript tests
npm test

# Run Python end-to-end tests
npm run test:e2e

# Run all tests
npm run test && npm run test:e2e

🔗 Integration Testing

# Test with Claude Code
# 1. Start the MCP server
npm start

# 2. In Claude Code, use any of the MCP tools to test integration
# 3. Check documentation output in your configured path

📊 Performance

The system is designed for efficient operation with:

• 📁 File Watching: Monitors project files for changes
• ⚖️ Queue Processing: Asynchronous task processing
• 💾 Storage: Organized markdown documentation files
• 🔄 Real-time Updates: Responds to file system events

🛡️ Security

🔒 Security Features

• 🛡️ Path Sanitization: All file paths are validated and sanitized
• 🔐 Process Isolation: Subprocess execution with limited permissions
• ✅ Input Validation: All user inputs are validated and escaped
• ⚠️ Error Handling: Comprehensive error handling prevents crashes
• 📏 Resource Limits: Memory and CPU usage limits prevent abuse

🎯 Best Practices

• Run with minimal required permissions
• Use dedicated user account for production deployments
• Monitor resource usage and set appropriate limits
• Regularly update dependencies for security patches

🤝 Contributing

We welcome contributions! Please see CONTRIBUTING.md for details.

🛠️ Development Setup

# Clone and setup
git clone https://github.com/Puneet8800/claude-auto-documenter-v2.git
cd claude-auto-documenter-v2
npm install
pip install -r requirements.txt

# Start development server
npm run dev

# Run tests
npm test

📏 Code Standards

• 🟢 JavaScript: ES2022+ with Node.js modules
• 🐍 Python: PEP 8 compliance with type hints
• 📚 Documentation: JSDoc and docstrings where applicable
• 🧪 Testing: Basic test coverage included

📄 License

This project is licensed under the MIT License - see the LICENSE file for details.

🙏 Acknowledgments

• 🎯 Anthropic: For Claude and the Model Context Protocol
• ⚡ Claude Code Team: For the excellent development environment
• 🌐 MCP Community: For standards and best practices
• 👥 Contributors: See CONTRIBUTORS.md for the full list

📞 Support

• 🐛 GitHub Issues: Report bugs or request features
• 📖 Documentation: Full documentation
• 💬 Community: Discussions

---

Made with ❤️ by the Claude Mechanics Master

Empowering developers with AI-assisted documentation automation