mcp-builder

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

• Third-party content exposure detected (high risk: 1.00). The SKILL.md explicitly requires fetching and reading public web content (e.g., "Use WebFetch to load: https://modelcontextprotocol.io/llms-full.txt", raw GitHub READMEs, and "use web search and the WebFetch tool as needed" to gather API documentation), so the agent will ingest untrusted third‑party web content that can directly influence tooling and subsequent actions.

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

• Potentially malicious external URL detected (high risk: 0.80). The skill instructs the agent at runtime to WebFetch external specification/docs (e.g., https://modelcontextprotocol.io/llms-full.txt and https://raw.githubusercontent.com/modelcontextprotocol/python-sdk/main/README.md) which would directly influence agent prompts/instructions and are treated as required reference material, so these runtime-fetched URLs present a risk of externally controlling agent behavior.