grill-me
Pass
Audited by Gen Agent Trust Hub on Apr 7, 2026
Risk Level: SAFE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to answer questions by exploring the codebase. This represents a potential surface for indirect prompt injection if files in the codebase contain instructions designed to manipulate the agent.
- Ingestion points: Local codebase files processed at runtime via file-reading tools.
- Boundary markers: None provided; the agent relies on its internal system instructions for safety.
- Capability inventory: The skill leverages standard file system read capabilities.
- Sanitization: No explicit sanitization of codebase content is defined within the skill.
Audit Metadata