azure-iot-edge

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

• Third-party content exposure detected (high risk: 0.70). The SKILL.md explicitly requires network access to fetch and read public Microsoft Learn documentation (via mcp_microsoftdocs:microsoft_docs_fetch or fetch_webpage) from the listed learn.microsoft.com URLs in the Category Index, so the agent ingests open/public third‑party content that can materially influence its decisions and tool use.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill explicitly requires runtime fetching of external documentation (e.g., https://learn.microsoft.com/en-us/azure/iot-edge/troubleshoot) via the mcp_microsoftdocs:microsoft_docs_fetch or fetch_webpage tool, and that fetched Markdown is injected into the agent context to guide/drive its responses, so external content can directly control prompts.