MCP Integration

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

• Third-party content exposure detected (high risk: 0.90). The skill explicitly instructs Claude Code to connect to arbitrary external MCP servers (SSE/HTTP/WS URLs like "https://mcp.asana.com/sse", "https://api.example.com/mcp", "wss://mcp.example.com/ws") and discover/use tools and returned data (see the SSE/HTTP/WebSocket sections, Lifecycle Management, and "Autonomous Agent" pattern), meaning the agent will ingest and interpret untrusted third-party content.

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

• Potentially malicious external URL detected (high risk: 0.70). The skill shows runtime MCP server endpoints (e.g., SSE example https://mcp.asana.com/sse) that Claude Code connects to to discover/register callable tools used by agents, which means the external server can execute actions or influence agent behavior at runtime.