dependency-track-skill

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

• Third-party content exposure detected (high risk: 0.80). The skill clearly ingests untrusted third‑party content at runtime — e.g., it downloads public resources (curl https://dependencytrack.org/docker-compose.yml, helm repo https://dependencytrack.github.io/helm-charts), mirrors and reads public vulnerability feeds (NVD, GitHub Advisories, OSS Index/Snyk/Trivy), and processes user-provided SBOMs via CI/CD scripts and API clients (GitHub Actions/Jenkins upload, upload-sbom.sh, python client), so the agent would read/interpret arbitrary external/user-generated content.