DocGen

A modular documentation generation system with support for academic paper implementation.

Overview

DocGen is designed to automate the creation and management of documentation for software projects. It supports a 2x2 matrix of project types:

• New Standard Projects: Creating documentation for new projects from scratch
• New Paper-Based Projects: Creating implementations based on academic papers
• Existing Standard Projects: Analyzing and documenting existing codebases
• Existing Paper-Based Projects: Adding or improving paper-based implementations in existing projects

Project Structure

src/
├── core/            # Core types and utilities
│   ├── types/       # Shared type definitions
│   └── utils/       # Common utility functions
├── modules/         # Reusable modules
│   ├── project-base/        # Base project functionality
│   ├── existing-analyzer/   # Analysis for existing projects
│   └── paper-architect/     # Paper processing and implementation
├── integrations/    # Integration layers for specific project types
│   ├── new-standard/        # New + Standard integration
│   ├── new-paper/           # New + Paper integration
│   ├── existing-standard/   # Existing + Standard integration
│   └── existing-paper/      # Existing + Paper integration
├── cli/             # Command-line interface
└── index.ts         # Main entry point for the library

Technology Stack

• Application Code: TypeScript (strict)
• Infrastructure: Python
• Documentation: Markdown with YAML frontmatter
• Dependencies: Node.js, npm

Installation

# Install dependencies
npm install

# Build the project
npm run build

# Link for development
npm link

Usage

DocGen can be used as a command-line tool:

# Create a new standard project
docgen new-standard my-project --description "My awesome project"

# Create a project based on an academic paper
docgen new-paper my-paper-impl --paper path/to/paper.pdf

# Analyze an existing project
docgen analyze path/to/project

# Analyze an existing paper-based project
docgen analyze-paper path/to/project --paper path/to/paper.pdf

# Auto-detect project type
docgen auto path/to/project

Development

Prerequisites

• Node.js (>= 14.0.0)
• npm (>= 6.0.0)
• Python (>= 3.8) for infrastructure scripts

Setup

# Clone the repository
git clone <repository-url>
cd DocGen

# Install dependencies
npm install

# Build the project
npm run build

# Run tests
npm test

Coding Guidelines

• All application code must be written in TypeScript
• Infrastructure and DevOps scripts are written in Python
• Follow the modular structure for adding new features
• Maintain separation of concerns between modules and integrations

Documentation

For detailed documentation, see the docs directory.

MCP System

DocGen includes a Model Context Protocol (MCP) system that enables AI-assisted development through tools like Windsurf and Claude Code.

Overview

The MCP system consists of:

• Docker-based MCP servers that provide various capabilities
• A TypeScript bridge for integration with Windsurf
• Automated deployment and configuration scripts

Requirements

• Docker and Docker Compose
• Node.js and npm (for building the TypeScript bridge)
• Python 3.6+ (for running deployment scripts)
• Windsurf (for AI-assisted development)

Setup and Deployment

1. Environment Configuration:

   Create a .env file in the project root with the following variables:

   GITHUB_TOKEN=your_github_token
   GITHUB_OWNER=your_github_username
   GITHUB_REPO=DocGen
   ANTHROPIC_API_KEY=your_anthropic_api_key
   

2. Deploy MCP Servers and Configure Windsurf:

   python scripts/deploy_mcp.py

   This script:

   • Builds and starts Docker containers for MCP servers
   • Builds the TypeScript bridge for Windsurf integration
   • Configures Windsurf to use the MCP servers

3. Restart Windsurf:

   After deployment, restart Windsurf to apply the configuration changes.

Usage

With Windsurf

Once the MCP system is set up and Windsurf is restarted, you can use Cascade AI within Windsurf to:

• Run tests and analyze results
• Manage GitHub issues
• Scan for TODOs in the codebase
• Execute the full "get-to-work" workflow

Simply ask Cascade AI to perform these tasks, and it will communicate with the MCP servers to execute them.

With Command Line

You can also use the command-line interface to interact with the MCP servers:

# Run the full get-to-work workflow
python scripts/get_to_work.py

# Run specific phases
python scripts/get_to_work.py --phase testing
python scripts/get_to_work.py --phase issues
python scripts/get_to_work.py --phase todos

Advanced Configuration

For advanced configuration options, see:

• MCP Implementation Plan - Detailed architecture and implementation details
• MCP Bridge README - Information about the Windsurf integration bridge
• Scripts README - Documentation for deployment and workflow scripts

Scripts

• Scripts for development are organized in:
  • scripts/windows/ - Windows-specific scripts (.ps1)
  • scripts/unix/ - Unix-specific scripts (.sh)
  • scripts/ - Cross-platform scripts

License

[License information]