Free tool · no sign-up
llms.txt checker and validator
Enter a site and this tool fetches its /llms.txt, checks it against the format published at llmstxt.org, tries up to 10 of its links, and scores it out of 100 with a fix for every problem. It also tells you whether the site serves an /llms-full.txt.
llms.txt checker
Validate any site's llms.txt
The format in one example
# Acme
> Acme makes invoicing software for freelancers in the US and EU.
Pricing starts at $9 a month. All plans include unlimited clients.
## Docs
- [Getting started](https://acme.example/docs/start.md): Set up your first invoice
- [API reference](https://acme.example/docs/api.md): REST endpoints and webhooks
## Optional
- [Changelog](https://acme.example/changelog.md): Release notes
The H1 names the site. The blockquote is the one-line answer to "what is this?". H2 sections group links, and a section named Optional marks links an agent can skip when it is short on context. The proposal also suggests serving markdown versions of pages (for example page.md), which is why the links above point to .md URLs.
How the score works
The checker starts at 100 and takes points off for each problem: 30 for an error (no H1, a missing or HTML file), 10 for a warning (no summary, no sections, relative links, broken links, a very large file) and 2 for minor notes. A missing file scores 0. The score is a quick read, not a ranking signal; fix the errors and warnings and the rest is optional.
Common problems
- Your app answers /llms.txt with HTML. Single-page apps and some site builders return the homepage or a 404 page with status 200. The checker spots this. Publish a static file at the root, or add a route that returns
text/plain. - The file is a sitemap in disguise. Hundreds of links with no notes help nobody. Keep it to the pages an assistant needs to answer questions about you, each with a one-line note.
- Links that redirect or 404. The checker tries up to 10 links. Fix broken ones first; they waste an agent's context.
Need to create one? The llms.txt generator for Shopify builds a file from a store's collections and policies, and the example above works for any site. To make sure AI crawlers can reach the rest of your site, run the robots.txt checker.
Questions
- What does a valid llms.txt need?
- Only one thing is required: an H1 with the site or project name. The format then expects a blockquote summary, optional free text, and H2 sections containing markdown link lists of the form "- [name](url): notes".
- Do ChatGPT, Claude or Google read llms.txt?
- llms.txt is a proposed convention, not a standard the AI engines have committed to. Coding agents and some AI tools read it when they are pointed at a site. It costs little to publish, but it does not replace a crawlable site, robots.txt access and good pages.
- Is llms-full.txt required?
- No. llms-full.txt (the full text of your key pages in one file) is a common convention, but it is not part of the llmstxt.org format. The checker reports it for information only.
- Why does the checker flag relative links?
- Tools often read llms.txt out of context, for example pasted into a prompt. Absolute URLs keep working there; relative ones don't.
- Can llms.txt live somewhere other than the root?
- Yes. The format allows a file in a subpath, such as /docs/llms.txt, to describe the pages under that path. This checker looks at the root file, /llms.txt.