Who it is forA developer automating explainer clips for many videos who wants the renderer without opening an editor.
The problemClicking through an editor panel for every clip does not scale.
The outcomeAn out/ folder with insert.mp4, preview.mp4, sfx_mix.wav, stills, plan.json and transcript.json.
What you need first
- Adobe Premiere Pro or DaVinci Resolve Studio
- pnpm and Node to build from source
Steps
- Clone github.com/erhanmeydan/kilnframe and run
pnpm installthenpnpm build - Export the speech segment you want as part.wav
- Run
node packages/core/dist/cli.js run --audio part.wav --style kil-lacivert --fps 30000/1001 --size 1920x1080 --out out/first - Open out/first and check insert.mp4, preview.mp4 and plan.json
- Run
pnpm checkafter any change (typecheck, lint, format and tests)
Prompt for your agent
Build kilnframe and render an insert from <wav> with style <style> at 1920x1080 into out/<name>. Confirm insert.mp4 and sfx_mix.wav exist and report the scene count from plan.json.
Caveat. You still connect an AI provider for planning; only the transcript, scene plan, style text and a few small stills are sent to it.
Use this in your AI agent
Ask your agent
Using the Deduped MCP, get the setup steps for erhanmeydan/kilnframe and "Render inserts from audio on the command line"
Add Deduped to Claude Code once
claude mcp add --transport http deduped https://dedupedai.com/mcp
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