🤖 TypeScript AI Documentation Generator

Automatically generate comprehensive JSDoc documentation for TypeScript code using local AI models via Ollama

✨ Features

• 🔒 100% Local - All processing happens on your machine, your code never leaves
• 🎯 Smart Parsing - Automatically detects functions, methods, parameters, and return types
• 📝 JSDoc Generation - Creates properly formatted documentation with descriptions
• 🚀 Powered by Qwen2.5-Coder - Optimized for code understanding and documentation
• ⚡ Fast Processing - Local LLM means no API rate limits or costs
• 🔍 Type-Aware - Understands TypeScript types, interfaces, and generics
• 🎨 Zero Configuration - Works out of the box with sensible defaults

📸 Demo

Before:

export function calculateArea(width: number, height: number): number {
  return width * height;
}

After:

/**
 * Calculates the area of a rectangle by multiplying its width and height.
 * 
 * @param {number} width - The width of the rectangle in units
 * @param {number} height - The height of the rectangle in units
 * @returns {number} The calculated area as width multiplied by height
 */
export function calculateArea(width: number, height: number): number {
  return width * height;
}

🚀 Quick Start

Prerequisites

• Node.js (v18 or higher)
• Ollama - Download here

Installation

# Clone the repository
git clone https://github.com/yourusername/ts-doc-generator.git
cd ts-doc-generator

# Install dependencies
npm install

# Install and start Ollama
# Download from https://ollama.ai, then:
ollama serve

# Pull Qwen2.5-Coder model (in a new terminal)
ollama pull qwen2.5-coder:7b

# Build the project
npm run build

# Try it out!
npm run doc examples/sample.ts

Basic Usage

# Generate documentation for a file
npm run doc path/to/your-file.ts

# Specify custom output location
npm run doc src/app.ts src/app.documented.ts

# Use with different Ollama model
npm run doc src/app.ts output.ts codellama

📖 Documentation

Command Line Options

npm run doc <input-file> [output-file] [model]

Argument	Required	Default	Description
input-file	✅ Yes	-	Path to TypeScript file to document
output-file	❌ No	<input>.documented.ts	Where to save documented code
model	❌ No	qwen2.5-coder:7b	Ollama model to use

Supported Code Constructs

• ✅ Function declarations
• ✅ Arrow functions
• ✅ Class methods
• ✅ Async functions
• ✅ Optional parameters
• ✅ Type annotations
• ✅ Generic types

Recommended Models

Model	Size	Speed	Quality	Best For
qwen2.5-coder:7b	4.7GB	Fast	Excellent	Code documentation (recommended)
qwen2.5-coder:14b	9GB	Medium	Outstanding	Complex codebases
codellama	3.8GB	Fast	Very Good	General code tasks
deepseek-coder	6.7GB	Medium	Excellent	Alternative to Qwen

🔧 Configuration

Change Default Model

Edit src/index.ts:

const model = args[2] || 'qwen2.5-coder:7b'; // Your preferred model

Adjust AI Temperature

Edit src/ollama.ts:

options: {
  temperature: 0.2,  // Lower = more consistent, Higher = more creative
  top_p: 0.9,
}

Customize Documentation Style

Edit the prompt in src/generator.ts to change:

• Documentation format
• Level of detail
• Included information
• Example generation

🎯 Examples

Document a Utility File

npm run doc src/utils/helpers.ts

Process Multiple Files

# Create a script to batch process
for file in src/**/*.ts; do
  npm run doc "$file"
done

Integrate with Build Process

Add to package.json:

{
  "scripts": {
    "predocs": "npm run build",
    "docs": "npm run doc src/index.ts docs/index.ts"
  }
}

🐛 Troubleshooting

Cannot connect to Ollama

# Make sure Ollama is running
ollama serve

# Verify in another terminal
ollama list

Model not found

# Pull the model
ollama pull qwen2.5-coder:7b

# List available models
ollama list

Slow generation

• Use a smaller model: qwen2.5-coder:1.5b or qwen2.5-coder:3b
• Ensure Ollama has enough RAM/VRAM
• Close other resource-intensive applications

Parser not finding functions

# Debug the parser
npm run test-parser examples/sample.ts

This will show the AST structure and what the parser detects.

Poor documentation quality

1. Try qwen2.5-coder:14b for better results
2. Adjust temperature in src/ollama.ts
3. Modify the prompt in src/generator.ts
4. Ensure your code has proper type annotations

Development Setup

# Clone your fork
git clone https://github.com/yourusername/ts-doc-generator.git
cd ts-doc-generator

# Install dependencies
npm install

# Make changes and test
npm run build
npm run doc examples/sample.ts

# Run the parser debugger
npm run test-parser examples/sample.ts

📄 License

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

⭐ Star History

If you find this project useful, please consider giving it a star! It helps others discover the project.

---

Made with ❤️ by [Pichutan]

Powered by Qwen2.5-Coder and Ollama