llms.txt is a proposed file, served at your root, that gives models a markdown index of your site: a short description and a list of your key pages. The idea is a projection made for machines, the way robots.txt is made for crawlers.
The honest take, because we would rather tell you straight: it is not a ratified standard, adoption is low, the major engines have not committed to honouring it and there is no measured evidence that having one lifts your citations. Worse, a hand-kept llms.txt drifts out of sync with the real site, which is a stale, conflicting projection, exactly the kind of mixed signal Kenovar is built to catch.
So Kenovar deliberately does not check for it, score it or generate one. If the standard matures and engines adopt it, we will revisit. Until then your effort is better spent on a correct sitemap, clean canonicals, server-rendered content and structured data, the projections engines actually read today.