bug-diagnosis
Pass
Audited by Gen Agent Trust Hub on Oct 22, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill establishes a methodical debugging framework, including building reproduction loops and testing hypotheses.
- [DATA_EXPOSURE]: The instructions explicitly include a security-positive requirement to 'Redact secrets before showing commands, logs, or captured requests' and recommend using environment variables for credentials to avoid exposure in command history or output.
- [REMOTE_CODE_EXECUTION]: While the skill involves running code (reproduction scripts, tests, and instrumentation), this is the inherent purpose of a debugging skill and is directed at the project's own codebase under user supervision. No untrusted third-party code downloads are specified.
- [EXTERNAL_DOWNLOADS]: The skill references a trusted external source (mattpocock/skills on GitHub) for its conceptual framework but does not execute remote code from it.
- [PRIVILEGE_ESCALATION]: No commands for privilege escalation (e.g., sudo, chmod 777) are present in the instructions.
Audit Metadata