Claude Optimization
Claude×
AI Can't Find Your Documentation

Claude: Make Your Documentation Discoverable by AI

AI Can't Find Your Documentation on Claude? Here's why it happens, the fix, and how to prevent it.

Claude Context

Make your website visible and citable by Claude, Anthropic's AI assistant known for nuanced understanding and detailed responses.

Allow Claude-SearchBot and Claude-User in robots.txt so Claude can find and read your pages

Focus on demonstrating genuine expertise (E-E-A-T signals)

Provide detailed, well-researched content with citations

Why This Problem Matters

Users ask AI for help with your product but it can't find or cite your documentation. Get your docs into AI knowledge.

Why This Happens in Claude

Documentation is only usable if Claude can reach it through Claude's web search, which Claude-SearchBot crawls for. Docs behind a login, on a blocked subdomain, or rendered client-side are invisible however good they are.

Signs You Have This Problem in Claude

1

Claude answers questions about your product without citing your docs

2

Claude gives generic answers instead of citing your docs

3

Users report AI couldn't help them with your product

4

Technical questions get wrong or incomplete answers

5

Competitor documentation gets cited instead of yours

How to Fix This in Claude

Follow these Claude-specific solutions:

1

Allow Claude-SearchBot in robots.txt, on your docs subdomain too

2

Serve docs as server-rendered HTML with one task per page

3

Make documentation publicly accessible without login

4

Publish docs as HTML pages, not just PDFs or apps

5

Implement proper heading hierarchy in docs

6

Link documentation prominently from your main site

7

Add TechArticle and HowTo schema to doc pages

Claude Technical Details

Anthropic documents three agents that all honour robots.txt: Claude-SearchBot for search results, ClaudeBot for training, and Claude-User for pages a user asks Claude to read.

Claude-SearchBot

Search / answers

Improves the quality of Claude's search results.

ClaudeBot

Model training

Collects public web content for model training.

Claude-User

User-requested fetch

Retrieves a page when a Claude user asks a question that needs it.

robots.txt rule to stay eligible for Claude answers:

User-agent: Claude-SearchBot
Allow: /

Source: support.claude.com

Prevent This Problem in Claude

Design documentation for both humans and AI

Test AI's ability to answer questions from your docs

Maintain clear documentation structure

Monitor documentation crawling in server logs

Need Help Fixing Claude Issues?

Get expert help resolving this and other Claude visibility problems.

Get Your AI Visibility Audit

Frequently Asked Questions

Why is this happening in Claude?

Documentation is only usable if Claude can reach it through Claude's web search, which Claude-SearchBot crawls for. Docs behind a login, on a blocked subdomain, or rendered client-side are invisible however good they are.

What is the first fix to prioritize in Claude?

Allow Claude-SearchBot in robots.txt, on your docs subdomain too