corey-ab-testing

Pass

Audited by Gen Agent Trust Hub on Jul 14, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADS
Full Analysis
  • [SAFE]: The skill provides instructional content, frameworks, and documentation templates for A/B testing. No suspicious behaviors, obfuscation, or persistence mechanisms were detected.
  • [EXTERNAL_DOWNLOADS]: The skill references established A/B testing tools and calculators (Evan Miller, Optimizely, VWO, AB Test Guide) and the author's public GitHub repository. These are reputable industry resources.
  • [PROMPT_INJECTION]: The skill establishes an attack surface for indirect prompt injection by directing the agent to read context from .agents/product-marketing.md. However, the skill lacks dangerous capabilities or instructions to perform high-privilege actions.
  • Ingestion points: .agents/product-marketing.md (SKILL.md)
  • Boundary markers: Absent
  • Capability inventory: No file writing, shell execution, or network exfiltration tools are used or instructed
  • Sanitization: Absent
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 14, 2026, 07:25 AM
Security Audit — agent-trust-hub — corey-ab-testing