Is your llms.txt written right?
Enter a domain. We fetch its /llms.txt and check it against the llmstxt.org proposal: the title, the summary, the sections and every link.
What the validator checks.
Results show here. Nothing is stored, and no AI model is involved: every check is a plain rule.
Found
/llms.txt answers with a text file, not a 404 or the site's catch-all HTML page.
Title first
The first line is one H1: the name of the site or project.
Summary
A blockquote under the title says what the site is, in a sentence or two. Optional, but it is what a model reads first.
Sections
Each ## heading holds a markdown list of links, written as - [Name](https://url): notes.
Links
Every link is absolute and well formed, so a model can open it from anywhere.
Optional
A section named Optional marks links a reader short on context may skip.
No file yet? Write one with the llms.txt generator, or read what llms.txt is first. An llms.txt helps only if crawlers can reach it, so check your robots.txt with the AI crawler checker too. Our own brief for models is on AI instructions, and the guide to AI visibility covers what moves AI answers more.
About llms.txt.
What is llms.txt?
A markdown file at the root of a site that tells language models what the site is and which pages matter, proposed at llmstxt.org in 2024.
Do AI engines read it?
Some agents and developer tools do. No major answer engine has said it ranks pages by it, so treat it as a cheap, clear brief for models, not a ranking lever.
What goes in llms-full.txt?
The full text of the pages llms.txt links to, in one file, for tools that want everything in a single fetch.
See what AI engines actually say about you.
Proofsource asks ChatGPT, Perplexity, Claude and Google the questions your buyers ask, and shows where you are named, where competitors are, and what to fix.