Setup & Installation
Or with the ClawHub CLI, for registry-managed skill folders outside a full OpenClaw workspace:
Version History
Fix display name
What This Skill Does
Sends voice message replies in iMessage that render as native inline voice bubbles with waveforms, not file attachments. Kokoro TTS generates audio locally, encoded with Apple's afconvert, then delivered via BlueBubbles. No cloud API calls are made for voice generation.
Uses Apple's native afconvert encoder to produce CAF/Opus audio identical to what Messages.app records, so recipients get a proper voice bubble with waveform rather than a file download.
When to use it
- Replying to a received voice message with your own voice message
- Sending a spoken response when typing is inconvenient
- Carrying on a voice-to-voice iMessage conversation through an agent
- Having the agent read a long reply aloud instead of as text
- Sending audio responses to group chats that use voice messages
Example Workflow
Here's how your AI assistant might use this skill in practice.
User asks: Reply to Marcus's voice message and tell him the meeting moved to 3pm
- 1Write the response text to a temp file to avoid shell metacharacter issues
- 2Run Kokoro TTS locally to generate a CAF audio file at the specified path
- 3Send the CAF via BlueBubbles with filename 'Audio Message.caf', contentType 'audio/x-caf', and asVoice true
- 4Send a companion text reply alongside for accessibility
Native iMessage voice bubble delivered to Marcus's number, playable inline with waveform display
Requirements
Accounts, API keys, or tools you or your AI assistant may need to set up while using this skill.