security-scanning

Pass

Audited by Gen Agent Trust Hub on Mar 22, 2026

Risk Level: SAFE
Full Analysis
  • [EXTERNAL_DOWNLOADS]: Installs security tools including Semgrep, OWASP dependency-check, Syft, Grype, and detect-secrets from official package registries and integrates official GitHub Actions for CI/CD workflows.
  • [COMMAND_EXECUTION]: Executes security scanning CLI tools using the Bash tool to evaluate project source code, container images, and dependencies for vulnerabilities.
  • [PROMPT_INJECTION]: The skill processes untrusted project data and manifest files, identifying an indirect prompt injection surface.
  • Ingestion points: Reads all files and configuration manifests in the project directory for scanning.
  • Boundary markers: No delimiters or boundary markers are used for the data being analyzed.
  • Capability inventory: Employs the Bash tool for running security scans and the Write tool for generating results and reports.
  • Sanitization: The skill relies on the internal parsing and validation mechanisms of the security tools themselves.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 22, 2026, 04:50 PM
Security Audit — agent-trust-hub — security-scanning