ab-test-setup
Pass
Audited by Gen Agent Trust Hub on Mar 11, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [SAFE]: No security threats detected. The skill is purely informational, providing guidance on A/B testing methodology, hypothesis construction, and sample size calculations. All external links point to well-known technology and A/B testing service providers.
- [NO_CODE]: The skill does not include any executable scripts (Python, JavaScript, shell) or configuration files that trigger automated actions. All logic is contained within markdown instructions and templates.
- [PROMPT_INJECTION]: The skill references a local context file (
.claude/product-marketing-context.md) to be read before interaction. While this is an ingestion point for untrusted data (Indirect Prompt Injection surface), the skill lacks any capabilities—such as file writing, network access, or command execution—that could be exploited via malicious content in that file. - Ingestion points:
.claude/product-marketing-context.md(referenced in SKILL.md) - Boundary markers: Absent
- Capability inventory: None detected (markdown-only skill)
- Sanitization: Absent
Audit Metadata