setup-agent-sign-in
Pass
Audited by Gen Agent Trust Hub on Nov 11, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill implements strong security practices for credential handling. It uses a local vault (
agent-browser auth) to store test user passwords, ensuring they are never passed as command-line arguments or stored in plaintext instructions. - [SAFE]: Data exfiltration is mitigated by redirecting sensitive command outputs (like
auth login) to/dev/nulland using--password-stdinto prevent passwords from appearing in process lists or shell history. - [SAFE]: The skill explicitly instructs agents to avoid methods that could expose passwords, such as reading HAR recordings or evaluating password field values.
- [SAFE]: Browser sessions are configured with fresh profiles and disabled password managers to prevent the UI from leaking credentials via 'save password' prompts.
Audit Metadata