How to generate llms.txt
- Enter the name of your site or project.
- Add a short summary that tells an agent what the site contains.
- Group the most useful pages under names such as Docs, API reference, Examples, or Optional.
- Give each link a clear name and, when useful, a short description.
- Preview the result, then copy it or download
llms.txt.
Choose the pages agents need
llms.txt is a guide, not a full sitemap. Link to the pages that help an agent answer questions or complete tasks. Put secondary material in an Optional group so an agent can skip it when it needs less context.
Prefer clean Markdown URLs
The current llms.txt proposal recommends links to agent-friendly content. Enter a page's .md or .html.md URL directly when one is available. The generated link keeps the same target.
Use the Markdown for Agents checker to test how a public page serves Markdown. Use the llms.txt validator to check a file that you changed or already published.
Publish the file yourself
Download the file and add it to your website. Use /llms.txt for a whole-site guide, or place it inside a path such as /docs/llms.txt when it covers only that part of the site. This tool does not crawl your site or update the file later.
Frequently asked questions
llms.txt is a small Markdown file that points AI agents to the most useful pages on a website. It gives the site or project name, a short summary, and groups of links with optional descriptions.
Publish it as llms.txt at the site root, such as https://example.com/llms.txt, when it describes the whole site. The current specification also allows a file inside a path, such as /docs/llms.txt, when it describes only pages below that path.
Yes, when clean Markdown versions are available. Agents can also use HTML pages. Direct .md links usually contain less navigation and page code. The generator keeps the same link target, including .md and .html.md paths.
Optional is a section name from the llms.txt convention. Put secondary links there so an agent can skip them when it needs a smaller context.
No. You choose every link. The tool does not fetch a sitemap, crawl pages, or update the file after you download it.
Yes. It uses the same parser as the llms.txt validator before it shows the file. You can also open the validator for line-by-line checks after you edit or publish the file.
No. Generation, preview, copy, and download run in your browser. The tool does not save your entries to an account or update your website.
What is llms.txt?
llms.txt is a small Markdown file that points AI agents to the most useful pages on a website. It gives the site or project name, a short summary, and groups of links with optional descriptions.
Where should I publish llms.txt?
Publish it as llms.txt at the site root, such as https://example.com/llms.txt, when it describes the whole site. The current specification also allows a file inside a path, such as /docs/llms.txt, when it describes only pages below that path.
Should the links point to Markdown pages?
Yes, when clean Markdown versions are available. Agents can also use HTML pages. Direct .md links usually contain less navigation and page code. The generator keeps the same link target, including .md and .html.md paths.
What does the Optional group mean?
Optional is a section name from the llms.txt convention. Put secondary links there so an agent can skip them when it needs a smaller context.
Does this tool crawl my site?
No. You choose every link. The tool does not fetch a sitemap, crawl pages, or update the file after you download it.
Does this tool validate the generated file?
Yes. It uses the same parser as the llms.txt validator before it shows the file. You can also open the validator for line-by-line checks after you edit or publish the file.
Is my content uploaded or saved?
No. Generation, preview, copy, and download run in your browser. The tool does not save your entries to an account or update your website.