Public-benefit foundation · UIC 208908769 · Sofia Transparency

Agent Readiness Auditor

Enter a URL. About 25 seconds of live checks against the standards AI assistants actually use, and an honest word about the ones they do not.

~25 seconds Nothing stored No email required

This tool needs JavaScript to stream results as each check completes. Every check, its point value and the scoring formula are published in full in the methodology, so you can run the same audit by hand.

Runs up to 20 polite requests from ISDI-AgentAudit/1.0 and respects your robots.txt. If you disallow us, we read only your robots.txt, llms.txt and sitemap, and show no score. Ten scans per hour per visitor.

What this scores, and what it refuses to score

Five categories are scored out of 100: access and crawlability, machine-readable content, structured data, agent affordances, and performance. A sixth — emerging standards such as DNS-AID records, MCP server cards and agentic-commerce hints — is probed and reported but deliberately worth zero.

That last decision is the point of the tool. Those standards are IETF drafts and proposals with no consumer implementations: no assistant consults them before reading or recommending a site. Scoring a site down for lacking them manufactures panic about work with no payoff today. We report them so you can watch them, and score you on what actually decides whether an agent can read you.

The full methodology, every check and every point value →