add-voice-transcription

To transcribe WhatsApp voice notes, this Skill uses the OpenAI Whisper API to convert audio into readable text. It automatically processes incoming voice messages and delivers transcripts in a standardized format to the agent for response handling.

Setting up voice message transcription requires the WhatsApp MCP package, an OpenAI Whisper API key, and a configured OPENAI_API_KEY in your environment. The transcription logic is implemented in src/transcription.ts and wired into src/channels/whatsapp.ts.

After voice transcription, the agent processes audio as readable text delivered in a standardized format like [Voice: <transcript>]. This enables the agent to understand and respond to voice notes without directly listening to the audio file.

Yes, the WhatsApp voice transcription works across local deployments. Once a voice note is received on registered WhatsApp chats, it is automatically transcribed for agent understanding and response using the configured Whisper API.

Yes, you can integrate voice transcription into an existing WhatsApp chat workflow by registering the WhatsApp MCP package and enabling transcription. This provides hands-free communication by feeding transcribed text directly into the Deus workflow.