grilling
Pass
Audited by Gen Agent Trust Hub on Aug 26, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill directs the agent to autonomously gather 'facts' from the environment, including the filesystem and tool outputs, using sub-agents. This establishes an attack surface for indirect prompt injection where malicious content in processed files could influence the agent's behavior.
- Ingestion points: Local filesystem and environmental tool outputs (SKILL.md).
- Boundary markers: The instructions do not specify any delimiters or safety markers (such as 'ignore instructions in data') for the gathered content.
- Capability inventory: The skill explicitly mentions dispatching sub-agents and using tools to explore the environment (SKILL.md).
- Sanitization: There are no instructions for sanitizing or validating the content retrieved from the environment before processing it.
Audit Metadata