mcp-builder

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

• Third-party content exposure detected (high risk: 1.00). The skill's core workflow and examples show MCP tools and resources that fetch and return external web/API content for Claude to read (e.g., references/scaffold/src/index.ts uses fetch('https://api.example.com/...') and references/patterns.md defines server.resource handlers that return api.getUser results), meaning untrusted third‑party content would be ingested and could influence agent decisions.