conversation-ab-testing

Pass

Audited by Gen Agent Trust Hub on Apr 3, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill content is purely educational and instructional, providing templates for building A/B testing systems. No prompt injections or malicious override patterns were identified.
  • [COMMAND_EXECUTION]: Python code snippets are provided for logical implementation of A/B tests (e.g., variant selection and statistical significance). These snippets do not include any dangerous system-level commands or unauthorized subprocess executions.
  • [EXTERNAL_DOWNLOADS]: The skill references the 'scipy' library for statistical calculations in its code examples. This is a standard and appropriate dependency for the described functionality.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 3, 2026, 05:45 PM