analytical-pm
Pass
Audited by Gen Agent Trust Hub on Oct 14, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill's research-first workflow involves fetching external data via web search and incorporating it into the final response, creating a surface for potential instruction injection from external sources.
- Ingestion points: User-provided questions and external benchmarks/data retrieved through 5-10 web searches (SKILL.md).
- Boundary markers: The instructions do not provide explicit delimiters for isolating retrieved search content from the agent's logic.
- Capability inventory: Ability to perform multiple web searches and generate detailed technical responses (~2000 words).
- Sanitization: No evidence of content sanitization or validation for retrieved search results is mentioned.
Audit Metadata