automated-testing

Pass

Audited by Gen Agent Trust Hub on Mar 18, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: No malicious patterns, obfuscation, or unauthorized access attempts were found.
  • [EXTERNAL_DOWNLOADS]: Configures installation of well-known security tools (garak, pyrit, safety, detect-secrets) through standard package managers.
  • [COMMAND_EXECUTION]: Executes Python scripts and security scanning tools within a controlled CI/CD environment.
  • [PROMPT_INJECTION]: The skill includes logic to detect prompt injection patterns in configuration files, rather than attempting to perform injections itself.
  • [CREDENTIALS_UNSAFE]: Appropriately handles sensitive information using CI/CD secret environment variables instead of hardcoded plaintext values.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 18, 2026, 07:13 AM