ab-testing
Pass
Audited by Gen Agent Trust Hub on Nov 4, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill is composed of informational guidelines and documentation templates. It does not include scripts, command execution, or sensitive data access.
- [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to ingest data from local project files like
.agents/product-marketing.md. Analysis shows no associated capabilities (e.g., network access, tool usage, or shell execution) that could be leveraged if these files contained malicious instructions. - Ingestion points:
.agents/product-marketing.md,.claude/product-marketing.md(specified inSKILL.md) - Boundary markers: None present
- Capability inventory: No tools or executable scripts are defined or used by this skill
- Sanitization: No sanitization logic is present for the ingested file content.
Audit Metadata