Fill in the fields and the file assembles itself. Everything runs in your browser, so nothing you type here leaves your machine.
Before you use it
It is a curated Markdown index that lives at the root of your domain and tells a language model what your site contains and what is worth reading. Jeremy Howard proposed it in 2024 and it is not an official standard.
Google does not use it. Worth saying first, because most of what gets written about this presents it as a recent obligation. What it is, is cheap: twenty minutes of work, and on some engines it changes what gets read from your site.
What decides whether it helps is size rather than anything technical. On a site under thirty pages a model crawls the whole thing unaided, so the file filters nothing.
Installing it
yourdomain.com/llms.txtrobots.txt does not block the root from AI crawlers,
because then nobody reads the file at allA file with relative links or empty sections uploads exactly the same and warns you about nothing.
That is why there is an llms.txt validator next to it: paste the file in and it says what is wrong before a model reads it.
And after this
It helps an engine understand the shape of your site. What decides a citation is having something worth citing, and that is content and original data rather than configuration.
If you want to know what state your site is in before changing anything, the diagnostic is the entry point, and it produces a list of priorities rather than a two-hundred-page report.