Setup & Installation
openclaw skills install @xadenryan/voice-wake-say
Or with the ClawHub CLI, for registry-managed skill folders outside a full OpenClaw workspace:
npx clawhub install voice-wake-say
What This Skill Does
Reads assistant responses aloud on macOS using the built-in `say` command. Activates only when the user message starts with the exact phrase "User talked via voice recognition". Strips markdown before speaking and falls back to text if `say` is unavailable.
Uses the local macOS `say` command instead of cloud TTS, so responses work offline with no API keys or external accounts.
When to use it
- Listening to assistant replies hands-free while working at a desk
- Getting verbal acknowledgment when kicking off long terminal commands by voice
- Dictating questions and hearing answers without looking at the screen
- Using voice input on macOS without sending audio to cloud services
- Confirming task completion during voice recognition sessions
Example Workflow
Here's how your AI assistant might use this skill in practice.
INPUT
User asks: "User talked via voice recognition on MacBook: summarize my last three tasks"
AGENT
- 1Detect that the message starts with the Voice Wake trigger phrase
- 2Run `say` immediately to acknowledge the request aloud
- 3Perform the requested task
- 4Strip markdown from the response text
- 5Run `say` again to speak the summary
OUTPUT
macOS speaks the acknowledgment and summary aloud; full text response also appears on screen