mcp-builder

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

• Third-party content exposure detected (high risk: 1.00). Yes — SKILL.md explicitly requires fetching public web pages (e.g., "Start with the sitemap at https://modelcontextprotocol.io/sitemap.xml" and "Use WebFetch to load https://raw.githubusercontent.com/.../README.md") and instructs using web search/WebFetch as part of the required Phase 1 workflow, so the agent is expected to read untrusted third‑party content that can materially influence tool design and subsequent actions.

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

• Potentially malicious external URL detected (high risk: 1.00). The evaluation harness connects at runtime to an external MCP server URL (e.g., https://example.com/mcp) to load tool definitions that are injected into the model as tools and to invoke those tools remotely (call_tool), which directly controls agent behavior and executes remote code.