scanning-iac-and-images-with-trivy

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (medium risk: 0.30). The runtime workflow in scripts/agent.py ingests outsider-authored free text indirectly when the user provides subject as a remote Git repository URL or an SBOM file/other external artifact, which the helper passes to trivy for JSON output parsing and gating.

MEDIUM W013: Attempt to modify system services in skill instructions.

• Attempt to modify system services in skill instructions detected (high risk: 1.00). The skill explicitly instructs running sudo commands that add apt keys/sources and install packages (modifying /usr/share/keyrings and /etc/apt/sources.list.d) and shows a docker run that mounts /var/run/docker.sock, all of which require elevated privileges and can modify or compromise host state.