What is a llms.txt?

The idea borrows its shape from robots.txt and sitemap.xml but solves a different problem. Those files tell crawlers what they may fetch and what exists. An llms.txt file tries to tell an AI system what actually matters: a short description of the site, followed by a linked list of the pages worth reading, often with a line of context each. Some sites also publish plain-text versions of key pages so a model can read them without wading through navigation, scripts and markup.

It is a proposal rather than a standard, and adoption by the major AI systems is uneven. That makes it a low-cost, low-guarantee addition. The realistic case for adding one is that it is cheap, it does no harm, and it forces you to decide which twenty pages on your site genuinely represent it. The case against treating it as important is that no AI system is obliged to read it, and none of it substitutes for having clear, crawlable content in the first place.

If you do write one, keep it honest and short. List the pages that answer real questions: what you do, what it costs, where you operate, how to get started. That is the same list that makes a website useful to humans, and the same list that makes an AI chat widget useful, since Clerkzo trains by crawling your site and answers visitors from what it finds there.