Agent Door - open your product to AI agents, one agent class at a time

Plan how your product opens itself to AI agents, provider side - which agent classes reach which surface, how each proves who it is and acts for a user, which path it gets and how agent usage is counted. Free in your browser: the agent class by surface matrix (training crawlers, AI search, user-triggered fetchers, browser agents, coding agents and fleets across marketing, docs, signup, the app and the API); your robots.txt read for 13 vendor tokens with real group and longest-match semantics (ClaudeBot is not Claude-User); the reachability gate from your CDN preset, signup CAPTCHA and probes; the agent-safe API checklist; the rungs - measurement, docs and Context7, agent credentials, skill plus CLI, MCP, provision-then-claim, discovery files - deleted and ranked by your answers; and a robots.txt that states the matrix. A metered run writes the identity and delegation plan, or the rollout. Built on the open-source skill @samber/agent-integration-surfaces, from samber/developer-platform-skills (MIT), by Samuel Berthe.