building-ai-agent-on-cloudflare

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

• Third-party content exposure detected (high risk: 0.85). The required runtime workflow ingests the user’s WebSocket message (onMessage(connection, message) → data.content / userMessage) and places it into the LLM messages array, which is outsider-authored free text from whoever connects to the agent (not chosen by the operating user).