How to Use AI to Document Code

AI tools can automatically generate comprehensive code documentation, from inline comments to complete README files. This eliminates the tedious manual work of writing documentation while ensuring consistency and accuracy across your codebase.

1. Choose your AI documentation tool. Install GitHub Copilot for VS Code, or use web-based tools like ChatGPT, Claude, or Codeium. For enterprise projects, consider GitHub Copilot for Business or Azure OpenAI Service. Each tool offers different strengths: Copilot excels at inline comments, while ChatGPT handles complex documentation structures.
2. Generate inline code comments. Place your cursor above a function or class and type `//` or `/**` depending on your language. AI will suggest comments explaining the code's purpose, parameters, and return values. Accept suggestions with Tab or modify them as needed. For existing code, select the function and use Ctrl+I (Copilot) or paste into ChatGPT with the prompt 'Add comprehensive comments to this code'.
3. Create function and class documentation. Select a function or class and prompt the AI with 'Document this function including parameters, return values, and usage examples'. The AI will generate detailed documentation following language conventions. For APIs, specify you want OpenAPI/Swagger format documentation.
4. Generate README files. Create a new README.md file and prompt AI with your project structure and main functionality. Use prompts like 'Create a README for a React project that handles user authentication and data visualization'. Include your package.json or requirements.txt file content for more accurate dependency documentation.
5. Document APIs and endpoints. Paste your API routes or controller code into AI and request OpenAPI specification generation. Use prompts like 'Generate OpenAPI 3.0 documentation for these REST endpoints'. The AI will create complete API documentation including request/response schemas, status codes, and examples.
6. Create code architecture documentation. Provide AI with your project's folder structure and key files. Request architecture documentation with prompts like 'Explain the architecture of this codebase and create a technical overview'. AI will generate diagrams, component relationships, and data flow explanations.
7. Review and refine AI-generated documentation. Read through all AI-generated documentation for accuracy and completeness. Verify code examples work correctly and update any outdated information. Use AI to expand sections that need more detail by prompting 'Expand this section with more technical details and examples'.