ab-testing-ecommerce

Pass

Audited by Gen Agent Trust Hub on Mar 16, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: No malicious patterns, prompt injections, or data exfiltration attempts were detected across the skill files.
  • [SAFE]: The skill recommends well-known and reputable experimentation tools such as GrowthBook, Convert.com, and Intelligems, and uses standard statistical methodologies.
  • [SAFE]: The provided evaluation criteria and task descriptions are designed to verify the functional correctness of experimentation logic, including deterministic bucketing via MD5 hashing and sample size calculations using standard z-scores.
  • [SAFE]: Implementation guidance for server-side variant assignment correctly prioritizes user privacy and legal compliance (e.g., price consistency) without introducing security vulnerabilities.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 16, 2026, 11:26 AM