llms.txt

llms.txt is a proposed standard for a Markdown file placed at the root of a website (/llms.txt) that gives large language models and AI agents a short, curated overview of the site: a title, a summary, and lists of the most useful links. An optional llms-full.txt can carry the full text.

Updated: 2026-10-16 · Glossary: AI search and agents

Format

The llms.txt proposal describes a file with an H1 title, a blockquote summary, optional free-text notes and H2 sections containing Markdown link lists, with an “Optional” section for links an agent can skip when context is short. Its purpose is to help models use a site at inference time, for example when an assistant or coding agent is asked about a product, rather than to control training or crawling.

What it does not do

• It is not a robots.txt replacement and grants or blocks nothing; crawler access is still set in robots.txt.
• Major search engines have not adopted it as a ranking or indexing input; treat it as a convenience for AI agents and IDE assistants.
• It cannot fix content an assistant cannot find or trust.

A fuller discussion of whether your site needs one is in llms.txt: what it is and whether your site needs one, and WordPress setup is in llms.txt on WordPress.

Why it matters for small sites

For a small brand, an AI assistant often has little to go on. A clear llms.txt that states what you do, what you do not do, prices and key pages costs an hour and reduces the chance of an assistant guessing wrong. More on AI visibility is on the AI search hub.

How Satelo uses it

satelo.io publishes /llms.txt with what Satelo is and is not, tools, pricing, the MCP server, topics and glossary, and /llms-full.txt with the full text of the key pages and blog posts, both generated from the same source as the website.

Related terms

• MCP (Model Context Protocol): MCP (Model Context Protocol) is an open protocol for connecting AI applications such as Claude, ChatGPT-based tools or Cursor to external tools and data.
• IndexNow: IndexNow is an open protocol that lets a website notify participating search engines the moment a URL is added, updated or deleted, instead of waiting for the next crawl.
• noindex: noindex is a rule, set in a robots meta tag or an X-Robots-Tag HTTP header, that tells search engines not to show a page in search results.

← All glossary terms