ab-test-setup

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFE
Full Analysis
  • [General Safety] (SAFE): The skill consists entirely of markdown-based instructions and procedural guidelines for A/B testing.
  • [No Code Execution] (SAFE): There are no scripts (Python, Node.js, Shell), system commands, or dynamic code execution patterns detected within the file.
  • [Data Security] (SAFE): No hardcoded credentials, sensitive file paths, or network exfiltration patterns were found.
  • [Prompt Integrity] (SAFE): The instructions are focused on process enforcement (hypothesis locking, metric definition) and do not contain adversarial injection patterns aimed at bypassing AI safety filters.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 06:35 PM