membership-push-targeted

Pass

Audited by Gen Agent Trust Hub on Sep 9, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill is designed with a secure human-in-the-loop architecture. It explicitly forbids automatic sends and requires the user to confirm a rendered draft before the classic_v3_create_push_notification tool is executed. Validation logic is included to prevent empty or placeholder messages, and the workflow follows a logical sequence of targeting resolution and verification. No malicious behaviors or security bypass patterns were found.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 9, 2026, 02:57 PM
Security Audit — agent-trust-hub — membership-push-targeted