grilling
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-provided plans and instructions the agent finds while exploring the environment (filesystem). It lacks explicit boundary markers or sanitization instructions for content read from these sources, which could allow instructions embedded in data to influence the agent's behavior.
- Ingestion points: Processes user input and external data from the filesystem (SKILL.md).
- Boundary markers: None present to distinguish user plans from agent instructions.
- Capability inventory: Implied access to filesystem reading tools to 'look up' facts.
- Sanitization: No sanitization or escaping of environment data is performed before processing.
Audit Metadata