synapse-client

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

• Third-party content exposure detected (high risk: 0.85). The skill ingests outsider-authored free text via serve()/subscribe() runtime NATS messages: any external agent can publish JSON envelopes whose payload (including arbitrary text fields) is decoded and passed into the user-provided handler/then potentially into the agent’s LLM context, creating an indirect prompt-injection path.