messaging-agents

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

• Third-party content exposure detected (high risk: 0.80). This skill sends messages to other agents who "can search the web" and "make API calls" (see "What the Target Agent Can and Cannot Do"), so the agent will receive and act on content originating from untrusted public web/APIs and that content could materially influence decisions or actions.