grill-me
Pass
Audited by Gen Agent Trust Hub on Jun 20, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is entirely instructional and guides the agent's conversational logic for design reviews.
- [PROMPT_INJECTION]: No evidence of system instruction overrides, safety filter bypasses, or jailbreak attempts.
- [DATA_EXFILTRATION]: No network-enabled tools or commands for transmitting data externally were detected.
- [REMOTE_CODE_EXECUTION]: The skill does not include any external package dependencies or remote script downloads.
- [SAFE]: Ingestion points: User-provided design plans (SKILL.md). Boundary markers: Not explicitly defined. Capability inventory: File system inspection (read-only) and user input tools. Sanitization: Not applicable as it uses standard agent dialogue.
Audit Metadata