Audio Transcription

Add OpenAI Whisper to transcribe audio files. Extract action items from voice memos, meetings, or any audio content automatically.

What You Can Do

Step 1: Install Whisper

Install Whisper and ffmpeg:

pip3 install openai-whisper
brew install ffmpeg

Step 2: Create Transcription Script

Create a script at ~/.openclaw/workspace/skills/transcribe.sh:

#!/bin/bash
export PATH="$PATH:/Users/$USER/Library/Python/3.9/bin"
whisper "$1" --model base --language English

chmod +x ~/.openclaw/workspace/skills/transcribe.sh

Step 3: Document in TOOLS.md

Add to your TOOLS.md file:

### Audio Transcription
- Tool: OpenAI Whisper
- Location: skills/transcribe.sh
- Usage: Send audio file, extract transcript and action items

Step 4: Use It

Send an audio file and ask:

• Transcribe this audio and extract action items
• Summarize this voice memo
• What are the key points from this meeting recording?