azure-ai-voicelive-py

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

• Third-party content exposure detected (medium risk: 0.30). The required runtime workflow continuously ingests outsider-authored free text via the user’s audio transcription events (e.g., conversation.item.input_audio_transcription.completed / event.transcript) and streams it back to the agent/model through the live WebSocket connection.