Free AI agent readiness checker
Is your website agent-ready?
Enter a domain and we request it the way ChatGPT, Claude, Perplexity and other AI agents do. You get a score, the raw findings behind it and a fix prompt you can paste into your coding agent.
Enter a website address to start the check.
- Free, no sign-up
- Results in under a minute
- Fix prompt for your coding agent
How the score works
107 scored checks in four groups.
Crawler rules and catalogs
Rules for named AI bots in robots.txt, and AI and agent catalog files with complete entries.
Pages and files
Text served without JavaScript, llms.txt, the sitemap, JSON-LD, pricing and trust pages, docs and OpenAPI, also on api., docs. and mcp. subdomains.
API behaviour
JSON error responses, OAuth metadata, pagination, rate-limit headers, versioning, and operations or page tools an agent could call.
Checkout
Agent payment protocols an automated buyer looks for: Universal Commerce, Agentic Commerce, AP2 mandates, x402 and Machine Payments. Most sites have none of these, and the section says so rather than counting it against them.
Reported, not scored
Emerging standards an agent may start using soon are checked and shown, but they never move the score up or down.
Fixed weights, not AI
A language model takes a second look at three checks - the AI policy in robots.txt, llms.txt quality and links to docs - and at what kind of site this is, so checks your site has no reason to implement are marked not applicable instead of failed. It never lowers a result, and the check says when it was involved.
Every result is a real request
Open a check to see what came back, then repeat the request yourself to confirm a fix.
No scripts, logins or forms
Content that only appears after JavaScript runs scores lower than it looks in a browser. Results reflect the site at the moment of the check.
Questions, answered
What the score tells you, and what it doesn’t.
What does agent readiness mean for a website?
Agent readiness is how easily AI crawlers and agents can reach your site, read its content and use what it exposes, such as documentation or an API. This checker scores it from 0 to 100 with 107 checks made over plain HTTP requests, and reports 20 more that cover emerging standards without scoring them.
Does AI decide the score?
No. Each check has a fixed weight and the score adds them up. A language model reviews three things that rigid rules misjudge: the AI crawler policy in robots.txt, how useful your llms.txt is, and which homepage links lead to documentation. It also reads what kind of site yours is, so checks it has no reason to implement come back as not applicable rather than failed. The model can never lower a result, and the check details say “per LLM review” when it was used.
Is this the same as an AI readiness assessment?
No. An AI readiness assessment usually asks whether a company, its data and its teams are prepared to adopt AI. This tool looks only at your public website and at what an AI agent gets back when it requests your pages and files.
Does it work as an llms.txt checker?
Yes. It looks for /llms.txt and /.well-known/llms.txt, checks that the file starts with a heading, contains Markdown links and stays under 30,000 characters, and follows those links to see whether they resolve.
Which AI crawlers does it test?
It requests your homepage as GPTBot, ChatGPT-User, ClaudeBot, PerplexityBot and DeepSeekBot, and reads your robots.txt rules for AI user agents, including OAI-SearchBot, Claude-SearchBot, Google-Extended and Applebot-Extended.
Will a higher score get my site cited by ChatGPT or Perplexity?
Not necessarily. The score shows whether agents can technically reach and read your site. Whether an AI answer mentions or cites you depends on much more, and no tool can guarantee it. To see what AI platforms actually say about your brand, use Glasa’s AI visibility tracking.
What is the fix prompt?
After a check, a language model writes a prompt for a coding agent - Claude Code, Cursor, Codex or another from the checks that failed. It lists the problems found on your site in order of impact, the change to make and how to verify it. Review the changes before you deploy them.
Is the agent readiness checker free?
Yes. Running a check and generating the fix prompt are free and need no account. Leaving your email for the report is optional.