grill-me

Pass

Audited by Gen Agent Trust Hub on May 9, 2026

Risk Level: SAFE
Full Analysis
  • [PROMPT_INJECTION]: No patterns of instruction override, safety bypass, or system prompt extraction were found. The instructions focus entirely on the intended design review task.
  • [DATA_EXFILTRATION]: No network-related commands (curl, wget) or attempts to read sensitive files (e.g., .ssh, .aws) were detected.
  • [EXTERNAL_DOWNLOADS]: The skill contains no references to external scripts, remote code, or third-party package installations.
  • [COMMAND_EXECUTION]: There are no shell commands, scripts, or system-level operations requested within the instructions.
  • [SAFE]: The instruction to explore the codebase is a standard capability of the underlying agent and is used here to inform the interview process rather than for malicious data gathering.
Audit Metadata
Risk Level
SAFE
Analyzed
May 9, 2026, 05:17 PM
Security Audit — agent-trust-hub — grill-me