Overview
Automatically generate structured, browsable documentation from any GitHub repository.
Paragon Wiki automatically generates comprehensive, browsable documentation from your GitHub repositories. Point it at a repo and get a full wiki with hierarchical topics, source file references, and an integrated chat for asking questions about the codebase.
What It Does
- Structured documentation — Automatically creates a multi-page wiki organized by topic from your repository's source code
- Hierarchical navigation — Browse topics in a sidebar tree with nested subtopics, just like a real wiki
- Source file references — Every wiki page links back to the source files it was generated from
- Integrated chat — Ask follow-up questions about the repository directly within the wiki viewer
- Keeps up to date — Regenerate the wiki at any time to reflect the latest code changes
How It Works
- Select a repository — Choose any connected GitHub repository from the wiki page
- Wiki generates — Paragon's agent clones the repo, analyzes the code, and builds a structured wiki
- Browse documentation — Navigate the generated wiki with a sidebar, table of contents, and search
- Ask questions — Use the integrated chat to ask anything about the codebase
Wiki Structure
Each generated wiki includes:
| Component | Description |
|---|---|
| Topic tree | Hierarchical sidebar navigation organized by theme (e.g., Architecture, API, Components) |
| Pages | Individual documentation pages with rendered Markdown, code blocks, and diagrams |
| Table of contents | Right-side TOC that tracks your scroll position within a page |
| Source files | Collapsible section on each page showing which files were referenced |
| Chat | Bottom panel for asking questions with full context of the current page |
Next Steps
Generating a Wiki
Create your first wiki from a GitHub repository
Browsing & Navigation
Navigate topics, pages, and the table of contents
Wiki Chat
Ask questions about your codebase within the wiki
Managing Wikis
Regenerate, delete, and organize your wikis