Score your site for ChatGPT, Claude, Perplexity and Google AI Overviews in under 10 seconds. 35+ signals tested across crawler access, discovery, schema, SSR, trust and frontier agentic standards (MCP, x402, OAuth, Web Bot Auth, NLWeb). No signup to see your score.
Works on any public URL. We test against the same crawler signals used by GPTBot, ClaudeBot, PerplexityBot and Google's AI systems.
We're checking your site against 20+ agent-readiness signals.
These are the new and proposed standards that AI agents are starting to look for: MCP servers, Agent Skills, Web Bot Auth, x402, ACP, OAuth discovery, DNS-AID and friends. Most sites support zero today, so we treat these as bonus signals — early adopters get the recognition, late adopters get a roadmap. Nothing here is included in your main score.
A 4-page PDF with your score, the exact code/markup to fix each failing check, and an estimated impact score per fix. Sent to your inbox in 60 seconds.
No spam. We use your email only to send the report and an optional one-off follow-up.
Agent readiness isn't a single score — it's the sum of crawler access, content discoverability, structured meaning, server-side rendering and trust signals. Here's the full list:
robots.txt rules for GPTBot, ClaudeBot, PerplexityBot, Google-Extended, CCBot, Applebot-Extended and 5 more.
The emerging standard for guiding LLMs to your best content. We check /llms.txt and /llms-full.txt.
Sitemap.xml, canonical URLs, internal link density and crawl-friendly URL structure.
Schema.org structured data: Organization, Article, FAQ, Product, BreadcrumbList, HowTo and others.
Proper use of main, article, nav, header, footer and heading hierarchy (one h1, no skipped levels).
Title, meta description, OG tags and Twitter cards — the metadata agents use to summarise your page.
Does meaningful content appear in the raw HTML response, or only after JavaScript executes?
Word count, alt-text coverage and link descriptiveness — signals that the page is substantive.
HTTPS, language attribute, author markup, dates and Organisation schema.
AI agents like ChatGPT, Claude, Perplexity and Google's AI Overviews navigate the web like a fast, careful reader. They fetch your HTML, parse structured data, follow internal links and decide whether to cite you. Agent-ready sites get cited — everyone else gets summarised into oblivion. This tool scores your site against the same signals these crawlers care about.
Lighthouse measures Performance, Accessibility, Best Practices and SEO for human users on browsers. Agent readiness measures crawler-specific signals: AI bot access in robots.txt, llms.txt, schema markup, server-side rendering quality, and semantic HTML for agent navigation. They're complementary, not competing.
GPTBot (OpenAI training), ChatGPT-User (in-chat browsing), OAI-SearchBot (ChatGPT Search), ClaudeBot & anthropic-ai (Anthropic), PerplexityBot, Google-Extended (Google AI training), Applebot-Extended, Bytespider (ByteDance/Doubao), Meta-ExternalAgent and CCBot (Common Crawl, which feeds many LLMs).
No. The tool only fetches publicly accessible URLs. Sites behind login, IP allowlists or aggressive bot-blocking (Cloudflare under attack mode, etc.) may return partial results or fail entirely. For protected sites, contact us for a manual audit.
We log the URL and the score for our own analytics (to improve the tool) but never share it. If you submit your email for the PDF report, that email is used only to send the report and an optional one-off follow-up. No third-party sharing, ever.