Do the links in my llms.txt file work?
An llms.txt file is a list of links for an AI to follow, so a link that leads nowhere is the file failing at its one job. We try a small sample of the links in yours. Only a page that is clearly gone counts against you; a slow answer or a block counts as not measured.
What it actually is
From the llms.txt the scan has already fetched, we take the links in its file lists, spread a small sample across the whole file, and ask each one whether it opens. We ask lightly first and ask again properly where a server does not answer the light request honestly. Only a page that is confirmed missing is counted as broken. A timeout, a rate limit, a login wall or a bot block says nothing about the link, so it is counted as not measured.
Why it matters
The llms.txt proposal describes a file whose entries an agent follows to reach the pages that matter. When an entry leads to a missing page the agent gets nothing, and a file whose links have drifted out of date tells a machine the site is not looked after. Links go stale quietly, because nobody visits the file, so a periodic check is the only way the problem surfaces.
What good looks like
Every link in the file opens the page it names, relative links resolve from the site root, and the file is regenerated or reviewed whenever pages move or are removed.
Being honest about it
We try a handful of links, not all of them, so a pass means the sample was sound and not that every link is. A site that blocks automated requests to its own pages will show up as not measured instead of passing. The check only exists for sites that have an llms.txt, which is a separate check. Because it is new it sits under measured, not scored: it is shown in your report and never moves your score.
Where most sites go wrong
A file generated once and never updated, whose links point at pages that were later renamed or removed. The other is relative links written for a different folder than the file sits in, which resolve to addresses that never existed.
On the web
State of the Web · November 2026In November 2026, 91% of the sites where it applies (1% of the sites we could read) passed the “The links in llms.txt resolve” check.
Not enough history yet.
Based on the sites Bot Appétit has scanned. Most of these sites come from lists we scan for research. Not a random sample of the web.
Not sure where yours stands?
A free scan checks this and everything else in about 20 seconds — no signup. Or if it's already flagged and you'd rather it were simply handled, we do that too.
Flagged for this one? We handle it. No pitch, just a plain-English chat.
Book a call