grill-me
Pass
Audited by Gen Agent Trust Hub on Jul 27, 2026
Risk Level: SAFE
Full Analysis
- [DATA_EXPOSURE]: The skill is designed to read and write specific project documentation files (CONTEXT.md, CONTEXT-MAP.md, and files under docs/adr/). This access is constrained to the local repository and is necessary for its architectural review purpose. No network-based exfiltration was detected.
- [PROMPT_INJECTION]: No malicious override patterns, role-play jailbreaks, or safety bypass attempts were detected in the instructions. The skill utilizes a specific trigger ($grill-me) and focuses on structured design questioning.
- [EXTERNAL_DOWNLOADS]: The skill does not perform any network requests, external script downloads, or package installations.
Audit Metadata