Testing Skills With Subagents
Pass
Audited by Gen Agent Trust Hub on Oct 29, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is entirely documentation-based and does not include any executable scripts or binary files. No code execution risks are present.
- [SAFE]: No network operations, external downloads, or remote code patterns were identified. All resources mentioned are internal to the skill's directory structure.
- [SAFE]: No sensitive file access or hardcoded credentials were found. Example paths used in scenarios are for illustrative purposes and do not target actual system configuration files.
- [SAFE]: The instructions and pressure scenarios are part of a testing methodology and do not attempt to bypass the agent's safety protocols or inject malicious behavior.
Audit Metadata