Check if your site supports MCP, A2A, WebMCP, and llms.txt before you deploy an agent against it. Get a score across 5 pillars, find the exact failure points, and see a live agent session replay via real headless Chrome.
WebMCP Declarative API
A2A Agent Card
llms.txt Quality
robots.txt AI Bots
OpenAPI / MCP Server
PSI Core Web Vitals
Running agentic readiness audit…
Fetching page HTML + headers
Probing agent discovery endpoints
Parsing robots.txt + llms.txt
Testing content negotiation
Running PageSpeed Insights (PSI)
Running browser accessibility scan
Computing pillar scores
–Score
–
–
–
Agent Compatibility Chain
"AI readiness is a compatibility relationship, not a page attribute." · Andrea Volpini, Agent Web Compatibility Test
Pillar Strength Radar
Pillar Scores
Issue Distribution
Core Web Vitals (PageSpeed Insights · Mobile)
WebMCP Maturity
How far along the agent-callable spectrum is this site?
Agent Runtime Compatibility
Which AI agent deployment patterns can actually use this site today, based on Andrea Volpini's Agent Web Compatibility Test framework
Your Agentic Score measures protocol readiness. AX Audit measures what agents see inside the page.
A site can score well on MCP, llms.txt, and schema, yet still be unnavigable by an AI agent if the DOM has broken ARIA, ambiguous buttons, or collapsed content that never exposes text. AX Audit catches the layer the Agentic Score doesn't see.