Log inGet started

Free llms.txt checker and agent-ready site check

Type your website's address. We run 16 agent-ready checks, including llms.txt and AI crawler rules, and show the fix for each one that fails.

The agent checks local business sites fail most

CheckLocal business sites failingLocusPilot sites failing
Markdown source endpoints>99%20%
llms.txt64%7%
Link headers51%6%
Structured data for agents29%3%
Machine-readable formats9%1%

Source: LocusPilot scan of 500 local business sites listed in LocusPilot directories and 121 live sites built with LocusPilot, run through this checker. Failing means the check came back failed; partly done is not counted. Median score: 50 out of 100 for the local business sites, 100 for the LocusPilot sites. Last checked: November 2026.

How to use it

  1. Type your site's address, for example yourbusiness.com, and press Check my site.
  2. Give it 5 to 15 seconds. It reads your robots.txt, llms.txt, home page and the agent discovery files.
  3. Open the failed checks first: each shows what we found and how to fix it. Each site's result is saved for 24 hours, so after a fix, run the check again once that time is up.

How to fix it on your platform

WordPress
  1. With Yoast SEO, open Yoast SEO, then Tools, then File editor. With Rank Math, open General Settings, then Edit robots.txt. Without either, upload a robots.txt file to your site's root folder.
  2. In Rank Math, turn on Advanced Mode first if you do not see that tab. Rank Math cannot edit a robots.txt file that already sits in your root folder, so delete that file first or edit it directly.
  3. Give each AI crawler you care about its own group: "User-agent: GPTBot" on one line, then "Allow: /" on the next. Use Disallow for any you want kept out. A crawler with its own group skips your User-agent: * rules, so copy any Disallow lines you need, such as /wp-admin/, into its group.
  4. Save and open yoursite.com/robots.txt to confirm the new lines are there.
Wix
  1. Open yoursite.com/robots.txt first. Wix's default file has a Disallow rule for PetalBot, which the checker counts as an AI crawler, so the robots check usually passes as it is.
  2. To name more crawlers, go to SEO & GEO in your dashboard, scroll to Tools and settings, click Robots.txt Editor, then View File. Add a group for each one, such as ClaudeBot or GPTBot, with an Allow or Disallow line under it, then save.
  3. A failed llms.txt check on Wix usually comes down to one of three things: the site is on a free plan, has no custom domain connected, or is hidden from search engines. Wix makes the file itself once all three are sorted.
  4. Wix does not let you add your own response headers (Wix Studio Enterprise gets six fixed security headers only), so the Link headers check cannot pass.
Squarespace
  1. Squarespace writes your robots.txt for you and does not let you edit it.
  2. Its one AI option, Block known artificial intelligence crawlers in the Crawlers settings, keeps every listed AI crawler out. Leave it off if you want AI tools to read your site.
  3. With the option off, your robots.txt already names GPTBot, ClaudeBot and other AI crawlers, so the robots check passes and there is nothing to change.
  4. If llms.txt failed, make the file with our llms.txt generator. Its page shows where to paste it on Squarespace 7.1.
Shopify
  1. In your Shopify admin, open Online Store, click the three-dot menu on your published theme and choose Edit code.
  2. Click Add a new template and choose robots, then Create template. The template starts with Shopify's own rules as Liquid code. Leave that code as it is, because Shopify Support cannot help with this file.
  3. Below the default rules, add your AI crawler groups, for example "User-agent: ClaudeBot" and "Allow: /", then save. A crawler with its own group skips the User-agent: * rules above it, so copy any Disallow lines you need, such as /cart and /checkout, into its group.
Static sites
  1. Put robots.txt in the folder you deploy, with a group for each AI crawler you want to name.
  2. On Netlify or Cloudflare Pages, add a file named _headers to the folder you deploy (public/ in Astro or Next.js). Put /* on the first line and, indented under it, Link: </sitemap.xml>; rel="sitemap". That fixes the Link headers check.
  3. For the Markdown check, publish a Markdown copy of your home page at /index.md. In the same _headers file, add /index.md and, indented under it, Content-Type: text/markdown; charset=utf-8.

Common questions

16 checks from the agent readiness section of The Website Specification by Joost de Valk and contributors. They cover llms.txt and llms-full.txt, a Markdown copy of your home page, AI crawler rules and Content-Signal lines in robots.txt, JSON-LD, feeds and sitemaps, Link headers, and newer discovery files such as MCP, A2A agent cards, agent skills, NLWeb and schemamap, plus a DNS record for agent identity. Web bot auth cannot be seen from outside your server, and stable URLs need months of URL history, not one scan, so those two show as check by hand.
No. It checks that /llms.txt answers with a file and that the file has a top-level heading (a line starting with "# ") or at least one Markdown link. A file with neither is marked partly done. It does not test each line against the llmstxt.org format, which asks for an H1 name, a one-line summary in a > quote, then ## sections where each line is a Markdown link to a page followed by a short note, so check yours against that. If the file is missing or partly done, the result links to our llms.txt generator, which writes one in that format.
No. The check passes when robots.txt names an AI crawler, such as GPTBot or ClaudeBot, on a User-agent line with at least one Allow or Disallow rule under it, whether the rule lets it in or keeps it out. Rules for every bot under User-agent: * only count as partly done, and no robots.txt at all is a fail. Whether to block is your call. A blocked crawler cannot read your pages for the AI product it feeds.
Few site builders serve a Markdown copy of a page. The checker takes your home page address after any redirects, adds .md (so / becomes /index.md), asks for it with the header Accept: text/markdown, and passes only when the reply is a 2xx with a Markdown or plain-text content type. A normal HTML page or a redirect does not count.
Each check is weighted by how strongly the specification asks for it: required counts 3, recommended 2 and optional 1. A pass earns the full weight and partly done earns half. Your score is the points you earned divided by the points possible, shown out of 100. The one required check, stable URLs, is a check by hand, so in practice the score comes from the recommended and optional checks. Checks that are not applicable, checked by hand or could not run are left out. Optional files you do not have, like an MCP endpoint or an A2A agent card, are not applicable rather than failed.
No. It only reads public files, and it follows your robots.txt before it reads your home page or its Markdown copy. In robots.txt it goes by the name LocusPilotFreeTools. If your rules keep it off the home page, or your robots.txt cannot be read, the checks that need the page show as could not check.

Free llms.txt generator

Type your site's address. We read your sitemap and pages, then write an llms.txt that links only to your own pages. Check it, then upload.

Free local business schema generator

Fill in your details and copy LocalBusiness JSON-LD for your home page. It runs in your browser, so nothing you type reaches us.

Local citation sites for US businesses

These are US sites where a local business can get listed, each marked free or paid from what the site's own pages say.

Checks based on The Website Specification by Joost de Valk and contributors, licensed under CC BY 4.0. LocusPilot is not affiliated with the project.