`Google-Safety`

Google's special-case crawlers page says the Google-Safety user agent ignores robots.txt rules. It handles abuse-specific crawling, such as malware discovery for publicly posted links on Google properties, and is unaffected by crawling preferences. The page does not print a separate robots.txt User-agent line. This check matches the whole token and still reports the RFC 9309 decision.

robots.txt token: Google-Safety
Operator: Google
Kind: safety
robots.txt: does_not_follow. The vendor says this crawler does not follow robots.txt rules.
HTTP user-agent: Google-Safety The page prints this string as the HTTP user agent and does not print a separate robots.txt User-agent line.
IP list: https://developers.google.com/static/crawling/ipranges/special-crawlers.json The page says special-case crawlers operate from a different range than the common crawlers, published in this file. The URL returned HTTP 200 on 2026-10-18, creationTime 2026-10-16. Botcard does not copy the prefixes. The page says the reverse DNS mask matches rate-limited-proxy-***-***-***-***.google.com. A user-agent string can be spoofed.
Source: List of Google's special-case crawlers. Retrieved 2026-10-18, page says updated 2026-09-17. CC BY 4.0 (Google Developers page, except as that page notes).

JSON

curl -s https://whosthatbot.com/v1/crawlers/google-safety

Check this token

tools/call with name check_robots is the same check at POST /mcp. The pasted file disallows Google-Safety on /docs and allows *. The call does not fetch a host.

curl -s https://whosthatbot.com/mcp -H 'content-type: application/json' -H 'accept: application/json, text/event-stream' -d '{"jsonrpc":"2.0","id":1,"method":"tools/call","params":{"name":"check_robots","arguments":{"robots_txt":"User-agent: Google-Safety\nDisallow: /\n\nUser-agent: *\nAllow: /\n","path":"/docs","ids":["google-safety"]}}}'

That call returns fetched_target false. This row is disallow via group Google-Safety, rule Disallow: /.