parcours-rag/module3

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

• Third-party content exposure detected (high risk: 0.90). This skill explicitly downloads and ingests public third‑party documents (see references/bootstrap-workspace.md: curl from raw.githubusercontent.com and the step that fetches PDFs from ANSSI via the manifest URL) and then parses those PDFs into chunks that the retrieve/generate pipeline (CP2–CP5) uses, so untrusted web-hosted content is read and can materially influence the agent's outputs and actions.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.90). Yes — the bootstrap procedure downloads and writes executable agent/source files at runtime from raw GitHub URLs (e.g. https://raw.githubusercontent.com/etalab-ia/parcours-rag/main/src/mastra/agents/chat-agent.ts), which are required for the workspace and will be executed / injected into the model context, so they directly control agent behavior.