llms.txt generator and validator
Check your site's llms.txt against the proposal, and get a better draft built from your homepage and sitemap.
01Validation
02Draft llms.txt
What it checks
llms.txt is a markdown file at the root of a site that lists its key pages for AI models. 69% of the 110 SaaS homepages we audited publish one, against 50% of Indian startups, where some files were written by Shopify rather than the brand.
- How the file is served
- Whether /llms.txt returns a text file at all, or your HTML 404 page with a 200 status, which is common on sites built as single-page apps. We also check the Content-Type, redirects to other sites, and whether /llms-full.txt exists.
- The format
- The llmstxt.org proposal asks for a "# Name" title first, an optional "> summary" line, then "## Section" headings with "- [name](url): notes" link lists, and allows an "## Optional" section for pages a model can skip. We check each part, and flag relative links, links to other sites, and files too long to work as an index.
- Who wrote it
- Shopify now publishes an "Agent Instructions" file at /llms.txt for every store. It explains how an AI agent can shop the store but never says what the brand is. We flag it, since it can look like the job is done when it is not.
- A better draft
- We read your homepage and XML sitemap, take your name and meta description, group your pages into sections by folder (/blog, /docs, /pricing), and title each link from your own link text or its URL. You can copy it or download it as a file.
Questions, answered
- Is it free?
- Yes. There is no signup. Each run fetches up to six files from your site (the homepage, llms.txt, llms-full.txt, robots.txt and the sitemap) and runs a set of rules on them. No AI model is involved.
- Do ChatGPT, Claude or Gemini read llms.txt?
- Not as far as anyone has confirmed. llms.txt is a proposed convention, published at llmstxt.org in 2024, not a standard, and no major AI product has said it reads the file when answering questions. Several documentation platforms generate one automatically, and it takes minutes to add, so we treat a missing file as something to improve, not a failure.
- How many sites publish one?
- More than you might expect. In our audit of 110 SaaS homepages, 69% published an llms.txt. Among 105 Indian startups we could check, 53 did (50%), but 7 of those were the default file Shopify writes for every store. Counting only files the companies wrote themselves, India was at 44%.
- What is the Shopify "Agent Instructions" file?
- Shopify stores now serve an auto-generated /llms.txt that starts with "# Agent Instructions" and points AI agents to Shopify's shopping tools. It is useful for agents placing orders, but it does not describe the brand or link its key pages. If you run a Shopify store, you can replace it with your own file.
- What is llms-full.txt?
- A companion file some sites add, mostly documentation sites, with the full text of every page in one markdown file. It is optional. llms.txt stays the short index; llms-full.txt holds the long form.
- Why are some link titles in the draft made from the URL?
- Opening every page to read its title would mean hundreds of requests to your site. We use the link text from your homepage where we find it, and turn the URL into a title everywhere else. Rename them, and add a short note after each link on what the page covers.
- Do you store the sites I check?
- We keep each result for ten minutes so repeat checks are instant. We do not keep a list of the sites people check.
Made by 4080 Studio. For the rest of the groundwork AI search needs, run the AI visibility checker, or get in touch.