debug-flow
Pass
Audited by Gen Agent Trust Hub on Aug 23, 2026
Risk Level: SAFE
Full Analysis
- [DATA_EXPOSURE]: The skill includes explicit security constraints for its logging practices. It directs the agent to 'Never log: secrets, tokens, passwords, raw auth headers; full PII by default' and to prefer logging data shapes over full request/response bodies. This proactively mitigates the risk of accidental credential or data exposure during debugging sessions.
- [PERSISTENCE_MECHANISMS]: The skill implements a 'hard gate' cleanup process (Step 8) requiring the removal of every temporary log, helper, flag, and file added during the session. This prevents debug artifacts from persisting in the codebase or environment.
- [SAFE]: No obfuscation, malicious prompt injection, or unauthorized remote code execution patterns were detected. The skill uses standard logging facilities and expects human-driven reproduction in a local environment.
Audit Metadata