ScreenProof

Working prototype · Early access · iOS & Android · Web coming soon

Catch layout bugs on every screen size before your users do.

ScreenProof runs your Maestro flows on emulators of different sizes, flags broken layouts with AI, writes a fix RFC for each defect, and re-runs the flows to verify the fix actually landed.

  • Tests the real app, running
  • Not isolated components
  • Closes the loop: detect → fix → verify

01 · How it works

One cycle. Four steps. Repeats until the screen is clean.

The ScreenProof cycle Four connected steps in a loop: Run flows, Detect defects, Generate fix RFC, Verify fix. The loop returns to Run. 01 Run flows 02 Detect 03 Fix RFC 04 Verify 360 · 390 · 430 · 768
  1. 01

    Run your flows on a device matrix

    Your existing Maestro flows run on several emulators at once: small phones, large phones, tablets. No new test syntax. No component harness.

  2. 02

    AI inspects every screen

    A vision model reviews the captured screens and flags layout defects: overflow, overlapping elements, truncated text, clipped elements, and breakage specific to small or large screens.

  3. 03

    One fix RFC per defect

    Each defect becomes a short RFC: the screen, the device size, what broke, and a proposed correction. Written for the engineer who will fix it.

  4. 04

    Re-run and verify

    After the fix lands, the same flows run again on the same devices. The defect is closed only when it no longer reproduces.

02 · What it catches

The layout defects that slip past unit tests and snapshot diffs.

03 · Demo

A full cycle, start to finish.

What a run looks like: four emulators, two defects, two RFCs, one clean re-run. Simulated here; the product is in early access.

04 · Early access

We're onboarding mobile teams one at a time.

ScreenProof is a working prototype. Join the waitlist and we'll reach out when there's a slot for your team. Tell us about your app in the reply.

No spam. One email when your slot opens. Unsubscribe by replying.