split-test-evaluator
Pass
Audited by Gen Agent Trust Hub on Nov 3, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill performs localized statistical analysis on user-provided data. It relies exclusively on the Python standard library (
json,math,sys,statistics.NormalDist). - [SAFE]: No network operations, file system access, or credential usage were detected.
- [SAFE]: The script
scripts/ab_test_analyze.pyimplements standard statistical formulas for A/B testing (Z-test, Chi-square, Wald confidence intervals) and includes proper input validation to prevent division-by-zero or domain errors.
Audit Metadata