DEDUPEDCredentials. The voice comes through the ElevenLabs connector, so you must connect your own ElevenLabs account in Claude settings and keep its access under your control.
Who it is forAn educator or documentation writer who wants a narrated animated explainer with the picture synced to the voice.
The problemSyncing narration to animated visuals line by line is tedious, and every re-take breaks the timing.
The outcomeA narrated MP4 where the scenes are timed to the words of the voice-over, with pace checked in the review.
What you need first
- Claude Code
- Node 18+
- Playwright with Chromium
- ffmpeg on your PATH
- Python with faster-whisper and the ElevenLabs connector (only for voice-overs)
Steps
- Connect the ElevenLabs connector in Claude (Settings, then Connectors) and install faster-whisper with Python.
- Start the skill with a narration request:
/animate a narrated 45s explainer of how DNS works, math style, 16:9, voice from ElevenLabs - Approve the script and the storyboard as the skill presents them.
- Generate or re-time the voice with the tool:
node $SKILL/tools/voice.mjs pieces/my-piece - Export and review the narrated piece:
node $SKILL/tools/export.mjs pieces/my-piece --share --formats 9:16,1:1,16:9thennode $SKILL/tools/review.mjs pieces/my-piece
Prompt for your agent
Using the animate skill in cth9191/animate, create a narrated 45-second explainer about <topic> with a voice-over. Expect voice.wav and voice.json in the piece folder and an exported MP4, and verify that the review reports narration pace and that every script line is timed to the picture.
Caveat. The README names faster-whisper for timing and the ElevenLabs connector for the voice, but it does not state which voices or languages are supported.
Every Monday: the 5 hottest repos and exactly how to use them. Free, one email a week, unsubscribe anytime.
Get them freeShare: X LinkedIn Hacker News Email Copy link
Third-party repos we don't own. AI-written use cases can be wrong. Disclaimer