Validator
llms.txt Validator
llms.txt is new enough that most published examples are wrong. Paste yours and this checks it line by line against the structure the spec actually asks for, and names what is missing.
- Checks the real structureNot just “is it markdown”
- Explains each failureWhat to add, and where
- Paste, never fetchNothing leaves the page
Check it
Paste it here
Result
- H1 title —
- Blockquote summary —
- H2 sections —
- Markdown links —
- Links with descriptions —
- First problem —
llms.txt tells an assistant where the good documentation is. It does not replace a sitemap and it does not affect search rankings.
What the result means
The description after each link is the point
An assistant chooses which page to read from that sentence. A bare list of URLs gives it nothing to choose with, so it guesses. This is the check most files fail.
Absolute URLs, always
The file may be read from a different origin than the one it describes. A relative link that resolves correctly in your browser resolves to nothing in a crawler.
It is a curation, not a sitemap
A sitemap lists everything. llms.txt should list the pages you actively want quoted. If it has three hundred entries you have written a sitemap with extra steps.
It does not affect search rankings
This is worth stating internally before someone expects it to. It changes what assistants can find easily, not where you rank in a search results page.
Questions
It is a community proposal that has been widely adopted rather than a ratified standard. The structure below is what the common implementations expect.
At the root of your domain, at /llms.txt, served as text/plain or text/markdown.
Adoption is growing and uneven. It costs very little to publish and it is the clearest signal available today about which of your pages are worth quoting.
Want this run against your whole site?
Tell us the page your assistant lives on and the questions you care about. A person runs them and sends back the raw answers.