open-auto-doc

open-auto-doc is an automated documentation generation platform that analyzes GitHub repositories and creates comprehensive documentation websites using Claude AI. It automatically discovers code structure, APIs, components, and data models by running intelligent agents through the Anthropic Claude Agent SDK, then renders the analysis into beautiful, maintainable documentation sites built with Fumadocs (Next.js + Tailwind).

The tool eliminates manual documentation effort by using AI to understand architecture patterns, generate diagrams, document endpoints and components, and produce getting-started guides. It supports incremental analysis (re-analyzing only changed files) and can be integrated into CI/CD pipelines for auto-updating documentation as code evolves.

Who is this for? Open-source project maintainers, development teams, library/framework authors, and developers who want to automatically generate and maintain API documentation without manual effort. It's designed for teams using GitHub who value keeping documentation synchronized with code changes and want professional documentation sites deployed automatically.

What It Does

AI-powered documentation generator that analyzes GitHub repositories and automatically creates beautiful, hosted documentation websites.

• Interactive Documentation Setup — One command to set up comprehensive documentation with GitHub OAuth authentication, repository selection, API key configuration, and full analysis pipeline in an interactive guided flow.
• AI-Powered Architecture Analysis — Claude AI agents automatically analyze and document project architecture, including tech stack identification, module breakdown, data flow diagrams, entry points, key patterns, and Mermaid architecture diagrams.
• REST API Documentation Generation — Automatically extracts and documents all REST API endpoints with HTTP methods, URL paths, parameters (path/query/header/body), request/response bodies, authentication requirements, and relationship diagrams.
• UI Component Documentation — Discovers and documents UI components across frameworks with component names, descriptions, prop definitions (type, required status, defaults), usage examples, and categories.
• Data Model & Schema Documentation — Extracts and documents data models with field definitions, types, constraints, relationships, and generates Entity-Relationship (ER) diagrams using Mermaid syntax.
• User-Facing Features Analysis — AI synthesizes a product-oriented overview identifying key capabilities, features categorized by type, and real-world use cases from the technical architecture and code.
• AI-Generated Getting Started Guides — Automatically creates step-by-step getting started guides with prerequisites, installation instructions, quick start examples, configuration options, and code examples extracted from actual project files.
• Multi-Repository Analysis & Cross-Service Documentation — Documents multiple repositories together, analyzing shared dependencies, identifying API contracts between services, mapping service relationships, and generating cross-repo integration diagrams.
• Language-Agnostic Code Analysis — Works with TypeScript, JavaScript, Python, Go, Rust, Java, Kotlin, Ruby, PHP, C#, Swift and more, using AI-powered understanding rather than language-specific parsers.
• Fumadocs-Based Documentation Site Generation — Scaffolds a fully functional Next.js documentation website using Fumadocs with built-in full-text search, dark mode toggle, responsive sidebar navigation, and production-ready deployment configuration.
• Vercel Deployment Integration — Creates a GitHub repository for documentation and pushes content for seamless Vercel auto-deployment, with automatic redeployment on documentation updates.
• GitHub Actions CI/CD Automation — Adds a GitHub Actions workflow to source repositories that automatically re-analyzes code and updates documentation on every push to main, enabling zero-maintenance documentation sync.
• Incremental Analysis & Caching — Re-analyzes only changed files since last run using commit SHA tracking and diff analysis, significantly reducing API costs and regeneration time for large repositories.
• Project Context Via CLAUDE.md — Reads optional CLAUDE.md files to inject insider knowledge into AI prompts, improving documentation accuracy through developer-provided context about conventions, architecture patterns, and project conventions.
• Secure Credential Management — Securely stores GitHub OAuth tokens and Anthropic API keys with restrictive file permissions (0600), provides login/logout commands for credential lifecycle management.

Tech Stack

• TypeScript
• Claude Agent SDK (Anthropic)
• Node.js
• Zod (TypeScript validation)
• Commander.js (CLI)
• Handlebars (templating)
• Fumadocs (documentation site)
• Next.js
• Tailwind CSS
• GitHub API (Octokit)
• ESM modules

Project Stats