feat: Implement core real-time interaction capabilities for Gemini API

[vizvasrj]

This commit introduces the fundamental features for real-time communication with the Gemini API, encompassing:

• Real-time audio streaming: Enables bidirectional audio interaction using microphone input and API responses.
• Real-time video streaming (camera and screen capture): Allows streaming video from either a webcam or the screen, configurable via the MODE environment variable. Includes image resizing and encoding for efficient transmission.
• Text-based prompt functionality: Implements the ability to send text prompts to the Gemini API via the command line.
• Graceful shutdown mechanism: Ensures proper cleanup of audio and video resources, including closing streams and websocket connections, upon application exit.
• Initial README documentation: Provides a comprehensive guide on project setup, configuration, and usage examples for various modes.

[vizvasrj]

any one going to review his?

[qiaodev]

Thank you for your contribution. This repository is not yet open for external contributions. We will update once we have established a process for accepting pull requests.