Shared GPT Live frontend

assistant.py owns the GPT Live WebSocket, voice session, incoming events, streamed caller audio, and connection lifecycle. Both delegation modes reuse this frontend. transport.py owns shared protocol validation, session payloads, authentication, and WebSocket connections.

Edit prompts/voice.txt to change the assistant's spoken behavior. Configure the GPT Live model, voice, endpoint, and credentials in the selected .env file or shell. See environment-file selection and security and artifact handling.