mcp-builder

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

• Third-party content exposure detected (high risk: 0.80). SKILL.md explicitly instructs the agent to fetch and read public web resources (e.g., "Use WebFetch to load: https://modelcontextprotocol.io/llms-full.txt" and raw GitHub README URLs) and to "use web search and the WebFetch tool as needed," meaning the agent will ingest open/public third‑party content (including user‑authored GitHub content) as part of its workflow, which 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: 0.80). The skill explicitly instructs using WebFetch at runtime to load external specification and SDK README files (https://modelcontextprotocol.io/llms-full.txt, https://raw.githubusercontent.com/modelcontextprotocol/python-sdk/main/README.md, https://raw.githubusercontent.com/modelcontextprotocol/typescript-sdk/main/README.md), which are fetched and used to guide the agent's behavior (i.e., injected into the model's context/instructions) and are required during the workflow.