confluent

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

• Third-party content exposure detected (high risk: 0.70). The skill instructs the agent to fetch and interact with external Confluent data via Membrane actions and the proxy (e.g., "membrane request CONNECTION_ID /path/to/endpoint" and actions like "get-topic"/"list-topics"), which exposes user-generated/untrusted Kafka topic content that the agent is expected to read and that could materially influence subsequent actions.