grill
Pass
Audited by Gen Agent Trust Hub on Sep 28, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze user-provided plans or designs, which serves as a potential surface for indirect prompt injection if the data contains malicious instructions.
- Ingestion points: Processes user plans, designs, and project context provided during the "grilling" interaction.
- Boundary markers: No explicit delimiters or instructions are provided to the agent to ignore embedded commands within the user's design documents.
- Capability inventory: The skill explicitly instructs the agent to "inspect codebase first," implying the use of file reading tools to access local project data.
- Sanitization: There are no defined mechanisms to sanitize or validate the content of the designs or the codebase before processing.
Audit Metadata