Agent Experience
How ready a site is for AI agents to read, discover, and operate on it
Agent Experience (AX) rules score how ready a site is for AI agents — assistants, answer engines, and autonomous browsers — to read, discover, and operate on it. AX is to agents what UX is to people and DX is to developers.
The category splits into two questions:
- Can an agent read it? — LLM parsability, page-type clarity, structured data.
- Can an agent discover & operate it? — which AI crawlers you allow, and the durable site profile that anchors the audit.
Rules
Reports which AI-agent crawlers (GPTBot, ClaudeBot, PerplexityBot, Google-Extended, …) robots.txt allows or blocks
Flags main content that only appears after JavaScript runs — invisible to agents that read raw HTML
Detects /llms.txt (+ /llms-full.txt) at the domain root and checks its basic Markdown format
Checks whether the site serves text/markdown via content negotiation or a .md variant
Resolves the per-domain site profile that gates cloud features and which rules apply
Analyzes how well LLMs can parse and understand the content
Checks the AI-classified page type agrees with declared structured data
Detects if content appears to be AI-generated (currently disabled)
Roadmap
More AX checks are planned: WebMCP / agent-accessibility parity with Lighthouse agentic browsing.
Disable All Agent Experience Rules
[rules]
disable = ["ax/*", "ai/*"]