Maintenance Automation

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

• Third-party content exposure detected (high risk: 0.70). The skill explicitly runs tools that ingest public package registry and advisory data (e.g., analyze_python_dependencies uses "pip list --outdated" and "pip-audit", analyze_node_dependencies uses "npm outdated" and "npm audit") and the code uses those external, untrusted outputs to generate recommendations and drive automated updates (subsequent pip/npm install), so third‑party content from public registries can materially influence agent actions.