Speech to Text

User: Transcribe this audio file: /path/to/recording.mp3
Agent: [uses stt tool] → "Here is the transcription..."

{
  "name": "stt",
  "description": "Transcribe audio to text using OpenAI Whisper",
  "input_schema": {
    "type": "object",
    "properties": {
      "file_path": {
        "type": "string",
        "description": "Path to the audio file to transcribe"
      },
      "language": {
        "type": "string",
        "description": "Language code (optional, auto-detected)"
      }
    },
    "required": ["file_path"]
  }
}