Turn an unfamiliar codebase into a browsable wiki with explanations and diagrams.
Ask natural-language questions about a repo and get answers grounded in the actual code.
Export the generated wiki as a standalone HTML file or an Obsidian vault.
Translate the generated documentation into another language like Chinese.
| porunc/codewiki | cybercal/hoic-baseline | hadriansecurity/openhack | |
|---|---|---|---|
| Stars | 39 | 39 | 39 |
| Language | Python | Python | Python |
| Setup difficulty | moderate | moderate | moderate |
| Complexity | 3/5 | 4/5 | 4/5 |
| Audience | developer | researcher | developer |
Figures from each repo's GitHub metadata at analysis time.
Self-hosted, needs an AI model configured through LiteLLM.
CodeWiki is a self-hosted platform that turns a code repository into a searchable, AI-generated wiki. The problem it solves is that large codebases are hard to understand for anyone new to them. There are no readable docs, just thousands of files. CodeWiki analyzes the structure of your code using a technique called AST parsing (which means it reads the grammar of code rather than just text), builds a graph of how everything connects, and then uses an AI model to write wiki pages that explain the codebase in human-readable prose with source citations and diagrams. You can then browse those pages in a built-in web interface, ask natural-language questions and get answers grounded in the actual code, export the wiki as a standalone HTML file or an Obsidian vault, and translate it into other languages like Chinese. It supports nine programming languages including Python, TypeScript, JavaScript, Java, Go, Rust, and C. The backend is built with FastAPI, the frontend with React and Vite, and the AI layer is handled through LiteLLM, which means you can plug in any compatible AI model. You would use it when onboarding to an unfamiliar project or when you want living documentation that updates as the code changes.
A self-hosted platform that turns a code repository into a searchable, AI-generated wiki with diagrams and source citations.
Mainly Python. The stack also includes Python, FastAPI, React.
Setup difficulty is rated moderate, with roughly 30min to a first successful run.
Mainly developer.
This repo across BitVibe Labs
Verify against the repo before relying on details.