Can AI systems use your website?

Each category receives a 0-100 score based on individual checkpoints. The overall score is a weighted average of all categories, converted to a letter grade from F (0-19) to A+ (95-100). Discovery carries 30% as an explicit versioned policy choice, not as a claim of measured causal impact.

Yes, IsAgentReady is completely free. Scan any website and get instant results with detailed scores, checkpoint-level pass/fail results, and actionable recommendations with code snippets.

llms.txt is a plain-text file at /llms.txt that provides a curated entry point to your site for AI agents. Unlike crawlers that scrape everything, agents can fetch specific pages for a task. llms.txt tells them which pages the publisher considers most relevant.

WebMCP (Web Model Context Protocol) and A2A (Agent-to-Agent) let AI agents interact directly with your website's functionality, not just read its content. WebMCP exposes actions through HTML form attributes or document.modelContext.registerTool(), while A2A Agent Cards describe your site's capabilities in a machine-readable format.

AI crawlers retrieve public pages at scale for uses such as search or training and depend on HTTP access controls like robots.txt. Browser agents operate a rendered page to complete tasks for users and may combine the DOM, accessibility tree, and screenshots. Server-rendered HTML gives both modes a dependable baseline, while browser testing is still needed for interactive flows.

Start with your scan's Quick Wins section, which shows the 3 changes with the highest point potential. Common quick wins include adding a robots.txt that allows AI bots, implementing JSON-LD structured data, and ensuring your pages render meaningful HTML server-side. Each recommendation includes copy-paste code snippets.

Yes. Install the IsAgentReady MCP server to scan websites directly from Claude Code, Cursor, Codex CLI, or any MCP-compatible editor. For fixing issues, install our open-source AI agent skills - they give your coding agent step-by-step playbooks for fixing what the scanner flags. Together they form a scan-fix-verify loop without leaving your editor.

Structured data such as JSON-LD with Schema.org vocabulary describes what a page represents: product, article, organization, or service. IsAgentReady measures whether that machine-readable signal is present and valid. The static scan does not measure or predict citation outcomes.