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 workflow explicitly directs the agent to fetch and ingest public third-party pages (e.g., "https://modelcontextprotocol.io/sitemap.xml" and raw.githubusercontent.com SDK READMEs) and to "use web search and WebFetch" during Phase 1, so untrusted web content can be read and could materially influence tool design and subsequent actions.