Does your site tell AI assistants what it contains?
llms.txt is an emerging, plain-text convention that describes a site and links its most important pages so AI tools can use it accurately. This checks whether yours exists, is well-formed, and points at real pages.
Fetches /llms.txt from outside your network and inspects its content. Saved to your account if you are signed in to a paid plan.
How this check works
Shipwork fetches /llms.txt from your site root and reports whether it exists, whether it has meaningful content, and how many pages it links to. It is optional and not yet widely honoured, but it is cheap to add and improves how assistants represent you.
What the llms.txt check examines
Shipwork requests /llms.txt at the site root from outside your network. It reports whether that request completed, whether an HTTP response was successful, whether the returned text has meaningful length, and how many absolute HTTP or HTTPS URLs it finds. A request failure is reported as an inconclusive check rather than proof that the file is absent.
This is a narrow public-file check. It does not parse whether every link is relevant, fetch every linked page, grade the writing, validate Markdown headings or determine whether an AI assistant has read the file. The check also does not test crawler access to your product pages. It shows the file that an external request can retrieve at the checked origin.
How to read each result
If the request cannot complete, Shipwork says it could not check the file on that run and recommends trying again. That can reflect a timeout or transient connection problem. If the response is not successful, the result says no llms.txt was found. If a file loads but contains fewer than 40 non-whitespace characters, it is flagged as nearly empty. A longer file gets a found result, with the number of absolute links counted when present.
A found result does not certify that the content is useful or that the URLs work. The count is based on URL-shaped text in the file; it is not a link crawl. Open a few links yourself, confirm they are public canonical pages, and check that the summary accurately describes your site. If you need to test whether bots can fetch pages, use AI access report or AI crawler check as appropriate.
Common llms.txt problems
A missing file often means it was uploaded to a subfolder, to the wrong domain, or only to a local theme preview. The conventional location is the origin root: https://example.com/llms.txt. Make sure the response is publicly available without a login and uses the same host visitors and crawlers should read. Check www and non-www separately if both serve content.
A nearly empty file gives an assistant little context. Add a short site description and a curated list of useful public pages with descriptive titles and summaries. Do not include private URLs, unpublished offers, claims you cannot support, or pages that return errors. Avoid stuffing it with every URL simply to increase the link count; a focused file is easier for a person to maintain and for a tool to navigate.
How to fix or create the file
Create a plain-text Markdown file named llms.txt with a top-level heading for the site, an optional one-sentence summary, and headings that group useful links. A simple entry can look like - [Shipping policy](https://example.com/pages/shipping): Delivery regions, rates, and estimated times.. Use real canonical URLs and write summaries that do not overstate what the linked page says.
Upload it to the root of the public site through your host, CMS, or deployment process. Then request the exact URL in a private browser window, check the returned file rather than an editor preview, and rerun Shipwork. The llms.txt generator can produce the core outline from your site name, short description and chosen URLs; review its output before publishing.
What this result means for AI visibility
llms.txt is an emerging convention, not a required search file. A successful check only confirms that the file was retrievable and passed this tool's limited content checks; it does not show that a named AI service uses it, that the service can crawl your pages, or that your site will be cited. Keep robots.txt, server access and page quality in view as separate issues.
Use the llms.txt guide for advice on selecting store pages, and use the AI crawler reference to distinguish bots that search, train or fetch on request. Treat any promise that a file alone guarantees AI citations as unsupported.
Questions
What is llms.txt?
Do I need one?
Where does it go?
How do I check if my website has an llms.txt file?
What does it mean if llms.txt could not be checked?
What should I put in an llms.txt file?
Does llms.txt make ChatGPT or other AI tools cite my site?
Why does the check find zero links in my llms.txt?
Keep reading
What this reads
- Reads public pages from outside your network, the way a crawler does.
- No signup, and nothing is stored on the free checks.
- One of 83 checks across discovery, content, markup, trust and catalogue data.
Cool, just ran an audit and got some insights! Let’s see if I can fix them :)
Whoah! this site is crazy, nothing to roast at all
cool as heck, tried it on my blog and the insights it gave are helpful and not bluff.