vertex-ai-api-dev

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

• Third-party content exposure detected (high risk: 0.90). The skill explicitly shows using Search Grounding and Url Context tools (references/structured_and_tools.md) and accepting arbitrary URLs/YouTube files (references/text_and_multimodal.md and SKILL.md TIP about search_documents/get_document), which fetch public third‑party content that the model is expected to read and use to shape responses, enabling indirect prompt injection.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill's MCP example launches an external MCP server at runtime with "npx -y @philschmid/weather-mcp", which will fetch and execute remote code (the npm package) and is used as a runtime tool that can influence model prompts and behavior, so it is a risky external dependency.