Real footage // No mockups // No timeline // 1080p60 // It frames the payoff // Works on any app // No API key to run // Open source · MIT // Real footage // No mockups // No timeline // 1080p60 // It frames the payoff // Works on any app // No API key to run // Open source · MIT //
NOW ROLLING · FREE & OPEN SOURCE

An AI director for your launch film.

Point supercut at your running web app. It reads the product, drives a real browser to perform the moments that sell it, and cuts a cinematic 60-second demo — real footage, edited automatically. No mockups. No timeline. No manual cuts.

Shot from a live app — zero manual editing ↓

REEL 01 — meridian.appNO CUTS ADDED BY HAND
REC 00:00:14:13
works on any app

Same director. Any product, any style.

supercut reads each app fresh — the copy, the colors, the money moments adapt per product. A light editorial dashboard or a dark analytics tool: the pipeline never changes.

A.Light dashboard type a record → frame the audit
B.Dark analytics search a service → frame the chart
why it's different

Your real product, shot like a launch film.

01

Real footage only

It drives your actual app in a real browser. Nothing is faked, mocked, or re-created.

02

It understands the product

Reads your routes and source, then crawls the DOM — so it films the money moments, not the landing page.

03

It frames the payoff

The camera holds on the result an action produces — the graph, the dashboard, the detail panel.

04

The Screen-Studio look

Spring zoom-to-cursor, motion blur, a padded backdrop, a clean 1080p60 export — automatically.

05

No API key to run it

The record + render path works fully offline. Only the AI director calls an LLM.

06

An open contract

The recorder writes a documented JSON event log. Any recorder can feed the renderer.

how it works

Five cuts. Each hands off plain JSON.

Stop at any frame, hand-edit the artifact, and resume the reel.

cut 01

Analyze

Read the source + crawl the app, pick the money moments. (LLM)

cut 02

Script

Write the filming recipe — schema-checked, no hallucinated selectors. (LLM)

cut 03

Record

A deterministic browser performs it, captured frame-by-frame.

cut 04

QC

Deterministic + optional vision checks, with bounded re-takes.

cut 05

Render

Cinematic compositing → final.mp4, ≤60s at 1080p60.

quick start

Point it at your running app.

zsh — supercut
# clone, build, and film your live app
git clone https://github.com/Co-Messi/supercut
cd supercut && npm install && npm run build

node dist/cli/index.js generate --url http://localhost:3000 --yes

No API key? The non-AI path runs standalone: record --recipe … then render --take …. Needs chromium + ffmpeg — run doctor to check.