grill-me
Pass
Audited by Gen Agent Trust Hub on Apr 22, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No malicious patterns, obfuscation, or unauthorized data access commands were detected in the skill instructions.
- [DATA_EXPOSURE_AND_EXFILTRATION]: The skill explicitly instructs the agent to explore the codebase and examine existing components. While this involves reading local files, it is restricted to the context of the user's project and is consistent with the skill's stated purpose of stress-testing designs within a development environment.
- [INDIRECT_PROMPT_INJECTION]: The skill creates an attack surface by instructing the agent to ingest data from the codebase. If files in the codebase contain malicious instructions (e.g., in comments), the agent might process them. However, there are no instructions here that encourage the agent to execute code or perform network operations based on that content.
- Ingestion points: Codebase files and project components.
- Boundary markers: None present.
- Capability inventory: File reading/exploration is the primary capability used.
- Sanitization: None present.
Audit Metadata