grilling
Pass
Audited by Gen Agent Trust Hub on Nov 3, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill creates a surface for indirect prompt injection by instructing the agent to automatically resolve questions by fetching data from the environment.
- Ingestion points: Data is ingested from the local filesystem and external tools when sub-agents are dispatched to find facts (SKILL.md).
- Boundary markers: No explicit delimiters or instructions are provided to help the agent distinguish between its own logic and instructions that might be embedded in the fetched data.
- Capability inventory: The skill utilizes sub-agents with the ability to access the filesystem and execute tools to gather information (SKILL.md).
- Sanitization: The instructions do not specify any validation, escaping, or sanitization of the environment data before it is integrated into the session context.
Audit Metadata