recipe-finding
Pass
Audited by Gen Agent Trust Hub on Oct 13, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data retrieved via web search and fetching tools.
- Ingestion points: Uses
WebSearchandWebFetchto collect recipe data from external websites (Phase 1). - Boundary markers: Lacks explicit delimiters or warnings for the agent to ignore instructions embedded within the fetched web content.
- Capability inventory: Employs
WebSearch,WebFetch, andMathtools; no sensitive file access or command execution capabilities are documented. - Sanitization: No procedures are defined for sanitizing or filtering the content retrieved from third-party sources before it is interpreted by the agent.
Audit Metadata