What it does
DocBeatBot is the crawler behind DocBeat, a free physician-finding site. It reads a health system's own public “Find a Doctor” pages and online-scheduling calendars to learn one thing: which physicians are accepting new patients. DocBeatthen sends patients to that system's own page or phone number to book.
It reads a few hundred pages per site, about once a week for appointment calendars and once every four weeks for directories, between two and ten seconds apart. It never reads competitors' sites or search engines, and it never collects anything about patients.
How it identifies itself
Every request carries the user agent Mozilla/5.0 (compatible; DocBeatBot/1.0; +https://docbeat.ai/bot) and comes from one fixed address, 66.175.209.80, whose reverse DNS is bot.docbeat.ai. The current list is at /bot/ips.json.
Requests are also signed. DocBeatBot uses Web Bot Auth (HTTP Message Signatures, RFC 9421) with an Ed25519 key whose public half is published at /.well-known/http-message-signatures-directory. A site, or a bot-management service in front of it, can verify that a request really came from us.
What it never does
It obeys robots.txt to the letter, including crawl-delay, on every host it touches. It never solves or works around a bot challenge: a challenge closes a site to it. It never books an appointment, submits a contact or appointment-request form, or enters personal data. It never uses another name, another address or a browser disguise. It does not run on sites that are not health systems' own.
How to keep it out, or let it in
Two lines in your robots.txt stop it entirely, and it will stop at once:
User-agent: DocBeatBot Disallow: /
A crawl-delay in the same group sets its pace. If your bot management blocks it and you would rather it read your pages, allow the user agent above or the address list, or write to hello@docbeat.ai. The same address reaches a person for any problem a request of ours caused; we answer within two business days.