building-ai-agent-on-cloudflare

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

• Third-party content exposure detected (high risk: 0.80). The skill's references (references/agent-patterns.md) explicitly show runtime fetching/ingesting of third-party content — e.g., ToolAgent.fetch('https://api.weather.com/...'), MCPClientAgent.addMcpServer('https://.../sse'), and RAG ingestDocument()/VECTORIZE queries — and those external results are fed back into AI calls and tool-decision flows, so untrusted third-party content can materially influence agent behavior.