Is your site AI agent ready?

AI agents are software systems that can take actions toward a goal with some autonomy—often by reading documentation, calling tools or APIs, or browsing the web. They range from chat assistants to crawlers and workflow bots. This site focuses on how clearly your public website signals rules and capabilities to those systems.

Rough buckets include: conversational assistants (chat over your help content), crawlers and indexers (search and AI training pipelines), and tool-using agents (MCP, plugins, or APIs that fetch pages on your behalf). Different agents respect different signals—robots.txt, llms.txt, and structured metadata all help reduce ambiguity.

It measures how well autonomous tools can discover your rules, preferred content formats (like markdown), and machine-readable capability files—without brittle full-site scraping.

Publish clear, machine-readable signals: a valid robots.txt (and optional Content-Signal lines), a sitemap, llms.txt if you want a curated LLM entry point, honest Content-Type headers, and well-known files such as MCP server cards or agent skill indexes where relevant. Fix broken redirects and avoid blocking helpful paths for legitimate automated clients.

It usually fetches a small set of URLs—homepage, robots.txt, sitemap, maybe llms.txt or /.well-known files—then decides what it may quote, summarize, or link to. It does not magically understand intent; it follows HTTP responses, headers, and the text you expose.

There is no universal certificate. Readiness means your public endpoints clearly state policy and capabilities and return consistent, honest responses. This scanner runs lightweight server checks against those endpoints and scores common gaps so you can prioritize fixes.

Common issues include unclear data or policies, brittle integrations, missing observability, and expectations that outpace what models can safely do. For websites specifically, vague robots rules, missing llms.txt, or misleading Content-Types make automated clients guess—and guesses are often wrong.

It is a simple, curated file models can read first so they route traffic respectfully and surface the URLs and policies you care about.

Optional directives in robots.txt that declare preferences for how AI systems use your content for training, snippets, and summarization.