ab-test-setup
Pass
Audited by Gen Agent Trust Hub on Mar 7, 2026
Risk Level: SAFEPROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [PROMPT_INJECTION]: The skill instructions include a step to read from .agents/product-marketing-context.md or .claude/product-marketing-context.md, which creates an indirect prompt injection surface where untrusted content could attempt to influence the agent's logic.
- Ingestion points: .agents/product-marketing-context.md and .claude/product-marketing-context.md (referenced in SKILL.md).
- Boundary markers: Absent.
- Capability inventory: The skill only performs text generation tasks for test design and lacks any capabilities for command execution, network requests, or filesystem modifications.
- Sanitization: Absent.
- [EXTERNAL_DOWNLOADS]: The skill provides links to several well-known industry resources and calculators, including Evan Miller's A/B tools, Optimizely, AB Test Guide, and VWO. These references are informative and point to trusted services for statistical analysis, involving no executable code downloads.
Audit Metadata