experiment-designer

Pass

Audited by Gen Agent Trust Hub on Apr 1, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill contains markdown documentation and a Python script for mathematical calculations. No malicious patterns such as prompt injection, data exfiltration, or obfuscation were detected.
  • [COMMAND_EXECUTION]: Includes a utility script scripts/sample_size_calculator.py used to estimate required sample sizes for A/B tests. The script uses standard Python 3 libraries (math, statistics, argparse) to perform calculations and does not interact with the network or the file system in an unsafe manner.
  • [DATA_EXPOSURE_AND_EXFILTRATION]: No sensitive file paths or network operations were identified. The script operates entirely on user-provided numerical inputs.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 1, 2026, 05:34 AM
Security Audit — agent-trust-hub — experiment-designer