adversarial

Pass

Audited by Gen Agent Trust Hub on May 3, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill is a legitimate security tool designed for codebase analysis and vulnerability detection.
  • [COMMAND_EXECUTION]: Utilizes local grep commands to identify risky code patterns such as eval, exec, and unsafe SQL queries during the reconnaissance phase.
  • [EXTERNAL_DOWNLOADS]: Mentions external tools like codex-cli and gemini-cli for cross-validation of security findings and attributes its strategies to the ZeroLeaks project on GitHub.
  • [DATA_EXFILTRATION]: Outlines strategies for scanning environment variables and detecting hardcoded secrets; these operations are performed locally as part of the intended security audit functionality.
Audit Metadata
Risk Level
SAFE
Analyzed
May 3, 2026, 11:06 AM