GravaBot and site controls
How Grava identifies its crawler, follows robots policy, and handles opt-out requests.
Grava is designed to be a recognizable guest on the public web.
User agent
Grava sends this user agent on its own navigation requests:
GravaBot/1.0 (+https://grava.dev/bot)Network and robots policy
Every outbound target must pass public-network validation before navigation and again after redirects. Grava respects robots.txt for the target path by default. A per-Recipe override is off by default and must be an explicit account decision.
Grava does not execute arbitrary customer code on its shared host. Recipes describe public URLs, selectors, fields, and Checks.
Opt out
Site owners can request exclusion by emailing hello@grava.dev with the hostname and enough information to verify control. Grava will review and apply the block at the service level.