Add voice chat extension

[Josh-XT]

Add voice chat extension

• Abstracts away the need to find the specific TTS provider for the agent on the front end.

Text-To-Speech providers and the order they're chosen

• Uses Streamlabs TTS with Brian voice by default.
• If USE_GTTS is true for the agent, it will take precedence over Streamlabs TTS
• If USE_HUGGINGFACE_TTS is true for the agent and it has a HUGGINGFACE_API_KEY, it will use the selected audio model on Hugging Face from the agent settings.
• If ELEVENLABS_API_KEY is not empty, ElevenLabs TTS will be used.

Python Usage Example

    import os
    from agixtsdk import AGiXTSDK

    ApiClient = AGiXTSDK(base_uri="http://localhost:7437", api_key="")
    # Get base64 audio from audio.wav and send it to the agent.
    with open(os.path.join(os.getcwd(), "audio.wav"), "rb") as audio_file:
        base64_audio = audio_file.read()

    ApiClient.execute_command(
        agent_name="gpt4free",
        command_name="Chat with Voice",
        command_args={
            "base64_audio": base64_audio,
            "conversation_results": 3,
            "context_results": 10,
        },
        conversation_name="Voice chat",
    )

TypeScript Usage Example

See the component that uses this in the NextJS front end.

https://github.com/AGiXT/nextjs/blob/main/components/systems/command/AudioRecorder.js