A read-only look at whether an agent can use your docs.
This scanner fetches a developer-tools company's public documentation the way an automated agent would, and keeps the raw responses. The point of a report is one specific fact — a status line, a robots rule, a page that is empty until JavaScript runs — not a grade.
It names itself
Every request uses this User-Agent, except one probe that compares a handful of well-known agent user-agents with a Chrome UA string. Only the User-Agent header changes. The scanner does not pretend to be a browser.
AgentReadinessScanner/0.1.0 (+https://agentready.rexstjohn.com/; read-only; docs and public metadata only)
The URL in that string is this page: https://agentready.rexstjohn.com/
What it will and will not do
It sends GET requests to public docs, robots.txt, sitemaps, and well-known metadata paths. It does not sign up, log in, guess passwords, fuzz hidden paths, or send anything that looks like an attack.
It stops. No domain receives more than 60 requests in a rolling 24 hours, across runs, serialized, with backoff. A Retry-After is honored. A 429 ends that domain's scan, and the unfinished checks are marked unknown rather than failed.
A check that times out, loops on redirects, or cannot be parsed is unknown. It is not recorded as a failure.
Reports are unlisted
A report is one company, one unguessable URL, and the raw evidence under each finding. There is no directory of who has been scanned and no single number that pretends to rank them. If you have a report, someone sent you the link. This scanner does not send email.