Voice Recorder (Whisper)

🚀 Voice Recording MCP Server

A Model Context Protocol (MCP) server for recording audio and transcribing it using OpenAI's Whisper model. It's designed to run as a custom extension for the Goose AI agent or as a standalone MCP server.

🚀 Quick Start

The Voice Recording MCP Server allows you to record audio and transcribe it using OpenAI's Whisper model. It can be used independently or integrated with the Goose AI agent.

✨ Features

• Record audio from the default microphone.
• Transcribe recordings using the Whisper model.
• Integrate with the Goose AI agent as a custom extension.
• Include prompts for common recording scenarios.

📦 Installation

# Install from source code
git clone https://github.com/DefiBax/voice-recorder-mcp.git
cd voice-recorder-mcp
pip install -e .

💻 Usage Examples

Basic Usage

Run as a Standalone MCP Server

# Run with default settings (base.en model)
voice-recorder-mcp

# Use a specific Whisper model
voice-recorder-mcp --model medium.en

# Adjust the sampling rate
voice-recorder-mcp --sample-rate 44100

Test with MCP Inspector

The MCP Inspector provides an interactive interface to test the server:

# Install the MCP Inspector
npm install -g @modelcontextprotocol/inspector

# Run the inspector and connect to the server
npx @modelcontextprotocol/inspector voice-recorder-mcp

Use with the Goose AI Agent

1. Open Goose and go to Settings > Extensions > Add > Command-line Extension.
2. Set the name to voice-recorder.
3. Enter the full path to the Voice Recorder MCP executable in the command field:

   /full/path/to/voice-recorder-mcp
   

   Or specify the model:

   /full/path/to/voice-recorder-mcp --model medium.en
   

   To find the path:

   which voice-recorder-mcp
   

4. No environment variables need to be set for basic functionality.
5. Start a conversation with Goose and introduce the recorder with the following instruction: "I want you to perform actions based on the transcribed text returned by the voice recorder. For example, when I dictate a calculation like 1+1, please return the result."

Advanced Usage

The server provides several useful tools:

• start_recording: Start recording audio from the default microphone.
• stop_and_transcribe: Stop recording and transcribe the audio to text.
• record_and_transcribe: Record audio for a specified duration and transcribe it.

📚 Documentation

Whisper Models

This extension supports multiple Whisper models:

Requirements

• Python 3.6+
• A hardware device with a microphone

Configuration

Configure the server's behavior through environment variables:

# Example configuration command:
export VOICE_RECORDER_API_KEY=your_api_key_here

Troubleshooting

Common issues and solutions:

⚠️ Important Note

• If you can't record audio, check microphone permissions and drivers.
• If there are transcription errors, ensure the API key is correct and the service is available.

Contribution Guidelines

Here's how you can contribute to the project:

1. Fork the repository to your personal account.
2. Create a new branch for feature development.
3. Submit your code and create a Pull Request.
4. Participate in issue discussions and bug fixes.

📄 License

This project is licensed under the MIT License. See the LICENSE file for details.