AIDotNet/OpenDeepWiki

OpenDeepWiki is the open-source version of the DeepWiki project, aiming to provide a powerful knowledge management and collaboration platform. The project is mainly developed using C# and TypeScript, supporting modular design, and is easy to expand and customize.

What it solves

Keeping project documentation up to date is tedious and often gets out of sync with the code. OpenDeepWiki automates this by ingesting a repository (from Git, a ZIP, or a local folder) and using AI to generate structured wiki content — READMEs, overviews, catalogs, detailed docs, and even mind maps. It then exposes that knowledge through a website, a chat assistant, and MCP (Model Context Protocol) endpoints so that humans and AI agents can query the repo's documentation.

How it works

  1. Import: You provide a Git URL, upload a ZIP, or point to an approved local directory. The system normalizes the source and prepares a workspace.
  2. Analysis & generation: It analyzes the tree and existing repo context, then uses configured LLM providers (e.g., OpenAI-compatible APIs) to generate README summaries, overviews, a wiki catalog, and document content. Translation into multiple languages (en, zh, ja, ko, es, fr, etc.) is optional and can run in parallel.
  3. Enrichment: Background workers handle follow-up tasks — translation, mind map generation (Mermaid), optional Graphify artifacts, and scheduled incremental updates when the repo changes.
  4. Serving: The generated knowledge is published on public routes like /{owner}/{repo}, exposed via a built-in chat UI, embeddable chat APIs, share links, and repository-scoped MCP endpoints (e.g., /api/mcp/{owner}/{repo}). Administrators manage users, roles, API keys, AI providers, models, skills, and more from an admin console.

The backend is ASP.NET Core (.NET 10) with background workers; the frontend is Next.js 16; data stores are SQLite or PostgreSQL; deployment is via Docker Compose.

Who it's for

  • Development teams that want auto-generated, up-to-date project documentation without manual effort.
  • Open-source maintainers who need public docs for their repositories.
  • Enterprises that want internal knowledge bases with role-based access, chat, and MCP integration for AI agents.
  • Organizations using multiple repositories who want a centralized wiki with multi-language support.

Highlights

  • Accepts Git URLs, ZIP uploads, or local directories as sources.
  • AI-generated READMEs, overviews, wiki catalogs, detailed content, translations, and mind maps.
  • Public SEO-friendly routes for each repo, plus a dedicated chat assistant and embeddable chat API.
  • MCP endpoints (repository-scoped or global) for AI tools to access repo knowledge.
  • Admin console for managing repositories, users, roles, API keys, AI providers/models, skills, and MCP providers.
  • Background workers for processing, translation, incremental updates, and Graphify artifact generation.
  • Optional support for Feishu, QQ, WeChat, and Slack incoming webhooks.
  • Docker Compose deployment with SQLite or PostgreSQL, plus a Makefile and Sealos option.

Related

  • Project
    AsyncFuncAI/deepwiki-open

    DeepWiki-Open is an automated tool that analyzes code repositories to generate interactive wikis, visual diagrams, and guided codemaps for documentation.

  • Project
    FSoft-AI4Code/CodeWiki

    An AI-powered tool that automatically generates and maintains hierarchical, dependency-aware documentation for large codebases, supporting incremental updates and artifact documentation.

  • Project
    ahmedkhaleel2004/gitdiagram

    GitDiagram is an AI-powered tool that automatically generates interactive architecture diagrams and narrated explainer videos for GitHub repositories, providing an MCP server for AI agents to query codebase structures.

  • Project
    xoai/sage-wiki

    A graph-based knowledge base that compiles raw documents into an interlinked wiki, enabling AI agents and humans to query a structured knowledge graph with citations.

  • Project
    arabold/docs-mcp-server

    Grounded Docs MCP Server is an open‑source, locally‑run service that scrapes and indexes documentation (websites, GitHub, npm, PyPI, local files) for any library version, stores it (optionally with vector embeddings), and exposes a MCP‑compatible HTTP endpoint. AI coding assistants (Claude, Copilot, Gemini, etc.) can query the server to get exact, up‑to‑date answers, reducing hallucinations while keeping data private.