rawdocs

Clean documentation from any public GitHub repo. Built for developers and their AI tools.

Try: facebook/react vercel/next.js openclaw/openclaw

How it works

Three steps. No signup. No API key.

Paste a repo

Enter any public GitHub repo URL or owner/repo shorthand.

Browse the tree

See every .md and .txt file in a navigable sidebar. No more hunting through GitHub's file browser.

Copy or fetch

One-click copy for pasting into ChatGPT or Claude. Or curl the URL for raw markdown.

Who it's for

🤖

Developers using AI assistants

Feed accurate library docs to Claude, ChatGPT, or Copilot. Stop the hallucinated API calls.

⚙️

AI agents and LLM crawlers

Request Accept: text/markdown and get clean markdown. No HTML to strip. Auto-generated llms.txt included.

🔍

Anyone exploring a repo

Discover documentation files you didn't know existed — hidden in .agents/, .github/, deep subdirectories.

Features

Zero indexing. Zero storage. Zero AI processing. Just a fast, thin viewer.

🔃 Content negotiation

Same URL, different formats. Browsers get rendered HTML with syntax highlighting. curl and LLMs get raw markdown. APIs get JSON.

📋 One-click copy

Copy raw markdown to clipboard instantly — no API call, no delay. Paste straight into your AI tool.

📄 llms.txt and llms-full.txt

Auto-generated per-repo index for LLM consumption. Full concatenated docs for single-file ingestion by large-context models.

🎨 Syntax highlighting

Shiki with GitHub Dark theme renders code blocks beautifully. The raw markdown stays clean — highlighting is for the HTML view only.

⚡ Fast by design

In-memory LRU cache, jsDelivr CDN for file content (SHA-pinned, immutable), Cloudflare edge caching. Sub-second responses.

🔭 AI-discoverable

JSON-LD structured data, x-markdown-tokens header, Content-Signal headers, and a robots.txt that welcomes every AI crawler.

Works with your existing tools

# Browse docs in your browser
open https://rawdocs.sivaram.dev/facebook/react

# Get raw markdown — just curl a .md URL
curl https://rawdocs.sivaram.dev/facebook/react/README.md

# Get all docs concatenated for LLM context
curl https://rawdocs.sivaram.dev/facebook/react/llms-full.txt

# Or use Accept header for content negotiation
curl -H "Accept: text/markdown" https://rawdocs.sivaram.dev/facebook/react/CONTRIBUTING.md

Try it now

Click any repo to see its documentation.