Implement core feature

[risu729]

This is a tool to transcribe a video file of a recorded interview.

The input is a Google Drive file URL passed through the Discord application command.
The output is uploaded to the same directory as the source file in Google Drive, and its link is returned to Discord.

This receives Discord application command through Gateway API, a WebSocket connection, to avoid opening ports of the server running on.

We extract audio files from the source video files and split them into files with a maximum size of 25 MB.
The file size limit is due to the limitation of Whisper API, which is the transcription AI for this.

Then, the transcription is fixed using GPT-4 or Gemini; both can be used depending on selection.

Referring to Google Drive, this downloads and uploads files through Google Drive API authenticated with a Google Cloud service account.
The directory where source files are uploaded must be shared with the service account beforehand.