How to allow, verify or block Google-Agent (Gemini tasks and user-triggered fetches)
Allow the Google-Agent token and the IP ranges in Google's user-triggered-agents.json.
Checked October 20, 2026. Our sources do not tie Google-Agent to Gemini Spark by name, and Google publishes no Spark-specific string.
How it identifies itself
| Signal | Value | Documented by the maker |
|---|---|---|
| user-agent | compatible; Google-Agent; +https://developers.google.com/crawling/docs/crawlers-fetchers/google-agent | Yes developers.google.com |
| ip-list | user-triggered-agents.json | Yes developers.google.com |
| web-bot-auth | Signature identity https://agent.bot.goog (experimental) | Yes developers.google.com |
- Allow it
- Allow the Google-Agent token and the IP ranges in Google's user-triggered-agents.json.
- Verify it
- Check the source IP against user-triggered-agents.json. Google is also testing Web Bot Auth under the identity https://agent.bot.goog, but says not every request may be signed yet.
- Block it
- Block the token or the published IP ranges at your firewall. Google says user-triggered fetchers generally ignore robots.txt, so robots.txt is advisory for Google-Agent.
What we can and can't confirm
Every signal above is published by the maker. Even so, a user agent string is easy to fake, so check requests against the maker's IP list or signature where one exists before you treat them as genuine.