Agent Readiness Checker
See whether agents can discover the site, read its rules and find real machine interfaces. The score changes with the kind of site you select.
One website. Only the checks that apply.
A law firm is not a failed API platform. An online store is not a documentation site. The selected profile controls the denominator.
Discovery
robots.txt, sitemap, llms.txt, Link headers and DNS-AID.
All profilesAccess and control
AI search crawlers, Content Signals and Markdown negotiation.
All profilesCapabilities
API catalogues, OAuth metadata, MCP, A2A, Agent Skills and WebMCP.
Apps and APIsCommerce
UCP, ACP, machine payments and x402 discovery signals.
EcommerceNew does not automatically mean essential.
- Established
- Full score weight
- Emerging
- 60% score weight
- Experimental
- 30% score weight
- Unable to verify
- Excluded, never failed
Can an AI understand and cite the page?
The AI Citability Analyzer maps answer passages, nearby evidence, entity clarity, provenance and freshness.
Open Citability AnalyzerQuestions
How is this different from the AI Citability Analyzer?
The AI Citability Analyzer maps one page's evidence and reusable answer passages. This tool checks site-wide discovery, crawler controls and optional agent protocols.
Does every website need MCP or OAuth?
No. Those checks only apply to the Application/API profile. A public content or business site is not penalised for lacking application protocols.
Do experimental standards affect the score?
Yes, but lightly. Established web fundamentals carry full weight, emerging checks carry reduced weight, and experimental protocols carry very little weight.
What happens when a request times out?
It is marked Unable to verify and excluded from the score. A timeout is never silently converted into a failure.