AX (Agent Accessibility) Audit checks how well your website supports AI agents. Just like web accessibility (a11y) ensures humans with disabilities can use your site, agent accessibility ensures AI agents can discover, understand, and interact with your services. The audit scans 7 signals: MCP discovery, structured data, API docs, machine-readable content, authentication, error handling, and rate limiting.

What is Agent Accessibility (AX)?

Agent Accessibility (AX) is a set of standards that make websites and APIs discoverable and usable by AI agents. As AI agents become the primary way users interact with services, sites that score high on AX will be preferred by AI assistants, search agents, and autonomous workflows. AX is to AI agents what SEO is to search engines.

How do I improve my AX score?

The AX Audit provides specific fix recommendations for each failed check, including code snippets you can copy-paste. Top priorities: add a .well-known/mcp.json file for MCP discovery, add JSON-LD structured data to your pages, create an llms.txt file describing your site for AI agents, and publish an OpenAPI spec for your API. AINative also offers professional AX optimization services.

Is the AX Audit free?

Yes, the AX Audit is completely free. Enter any URL and get a full score breakdown across 7 checks. Sign in with GitHub to unlock detailed fix recommendations with code snippets.

What is MCP discovery?

MCP (Model Context Protocol) discovery lets AI agents automatically find what tools and capabilities your API offers. By placing a .well-known/mcp.json file on your domain, agents can discover your endpoints, understand input/output schemas, and start using your API without human intervention. It is the single highest-impact AX signal.