GI AI Security public discovery is limited to publicly reachable metadata, browser-visible content and transparency observations. It does not authorize credential use, exploit execution, destructive actions or bypassing access controls.
Automated requests use an identifiable GI user agent and publish a security contact. Operators can request exclusion or clarification via security@general-informatics.com.
Active or red-team testing is a separate workflow and requires explicit authorization, a bounded scope, expiry, request/cost limits and revocation checks before execution.