product-manager
Pass
Audited by Gen Agent Trust Hub on Oct 4, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill's workflow includes ingesting potentially untrusted external data such as user feedback, business context, and stakeholder requests (SKILL.md, Phase 1). This data is processed and used to generate requirements and PRDs, creating a surface for indirect prompt injection. \n- Ingestion points: Business context, user feedback, and stakeholder requests are read during the 'Understand' phase. \n- Boundary markers: The skill does not define delimiters or markers to isolate external data from the agent's instructions. \n- Capability inventory: The skill is configured with access to tools including Read, Grep, Glob, and Bash (SKILL.md frontmatter). \n- Sanitization: No instructions are provided for sanitizing or filtering input data before interpolation into the agent's workflow.
Audit Metadata