experiment-design
Pass
Audited by Gen Agent Trust Hub on Aug 1, 2026
Risk Level: SAFEDATA_EXFILTRATIONCOMMAND_EXECUTION
Full Analysis
- [DATA_EXFILTRATION]: The skill exposes sensitive local environment information through the inclusion of a hardcoded absolute path.
- File: SKILL.md
- Evidence:
source /Users/zhangmingda/clawd/.venv/bin/activate - Description: The use of an absolute path reveals a local username ('zhangmingda') and specific directory structure, which constitutes information exposure of the host environment.
- [COMMAND_EXECUTION]: The skill instructs the agent to execute a Python script using a shell heredoc pattern.
- File: SKILL.md
- Evidence:
python3 << 'EOF'...EOF - Description: While the mathematical functions for power analysis are benign, the execution relies on activating a local virtual environment that is environment-specific and may fail or cause unexpected behavior on other systems.
Audit Metadata