deep-dive
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from external websites through web searches performed by sub-agents. This introduces a vulnerability surface where malicious instructions embedded in third-party web content could attempt to influence the agent's research synthesis or subsequent planning steps.
Audit Metadata