AI & agents
llms.txt Generator
Generate a practical llms.txt draft for AI assistants, agents, and documentation discovery.
How to use this tool
Enter your site name, canonical URL, short summary, documentation links, important workflows, and guidance for AI assistants.
Review the generated draft and publish it as `/llms.txt` when it matches your public site policy.
Keep the file short, canonical, and focused on pages that are useful for AI assistants, agents, and documentation discovery.
Why llms.txt can help AI agents
An llms.txt file gives assistants a concise, human-readable map of a site: what the project does, which docs matter, and what rules should be followed.
It is especially useful for documentation sites, developer tools, API portals, product knowledge bases, and agent-friendly websites.
Treat it as a helpful guide, not as access control. Use authentication, robots.txt, and server rules for enforcement.
What to include
Include canonical documentation, API references, tool pages, policy pages, and short notes about how AI systems should interpret the site.
Avoid dumping every URL. Link to stable indexes and pages that explain concepts clearly.
Do not include secrets, private endpoints, staging URLs, admin paths, or instructions that conflict with your public policy.
AI-agent safety notes
Tell agents which pages are documentation versus interactive tools.
Clarify whether tools run locally or require a server request.
Explain any restrictions around user data, credentials, uploads, and destructive actions.
Examples
Minimal llms.txt
Output# Example Docs > Developer documentation for Example API. Website: https://example.com ## Core documentation - https://example.com/docs - Documentation index - https://example.com/api - API reference
Agent guidance
Use guidance lines to tell assistants to prefer canonical docs, avoid private inputs, and respect authentication boundaries.
FAQ
Is llms.txt an official access control mechanism?
No. Treat it as a guidance file for AI assistants, not as a security boundary or replacement for authentication and authorization.
Where should I publish llms.txt?
The common practical location is the site root, for example `https://example.com/llms.txt`.
Should I list every page?
Usually no. List stable indexes, documentation, API references, and high-value pages instead of every URL.
Is this generated on the server?
No. The draft is generated locally in your browser and the inputs are not stored for analytics.