mcp-whatsapp-whisper

GBurgardt/mcp-whatsapp-whisper

3.1

If you are the rightful owner of mcp-whatsapp-whisper and would like to certify it and/or have it hosted online, please leave a comment on the right or send an email to dayong@mcphub.com.

MCP server that transcribes WhatsApp audio messages using local Whisper.

Tools
3
Resources
0
Prompts
0

mcp-whatsapp-whisper

Model Context Protocol (MCP) server that surfaces WhatsApp audio messages and transcribes them using OpenAI's Whisper API.

Requirements

  • macOS with WhatsApp Desktop (the project reads the local WhatsApp SQLite database and media files)
  • Node.js 18+
  • ffmpeg (brew install ffmpeg) for .opus → .mp3 conversion when needed
  • OpenAI API key available as the environment variable OPENAI_API_KEY

Install

npm install
npm run build

Environment

Set the OpenAI API key before starting the server. For example:

export OPENAI_API_KEY="sk-your-openai-key"

Run with PM2

export OPENAI_API_KEY="sk-your-openai-key"
pm2 start ecosystem.config.cjs

Cursor Config

Add to your Cursor settings (~/.cursor/mcp.json):

{
  "mcpServers": {
    "whatsapp": {
      "command": "node",
      "args": ["/path/to/mcp-whatsapp-whisper/dist/server.js"]
    }
  }
}

Tools

  • getRecentAudio: Get recent audio files from a contact
  • searchAudios: Search audio files by contact name or date
  • transcribeAudio: Transcribe audio through OpenAI Whisper (whisper-1)

License

MIT