gitmyhub

app-store-screenshots

TypeScript ★ 5.9k updated 21d ago

end to end app store screenshot creation using AI

This repository is a skill for AI coding agents such as Claude Code, Cursor, and Windsurf that generates production-ready screenshots for the Apple App Store and Google Play Store. Instead of a standalone app, it is a set of instructions and templates that your AI assistant reads and follows when you ask it to create marketing screenshots for a mobile app.

When activated, the skill asks you a series of questions about your app: what it does, its brand colors, the visual style you want, how many slides you need, and whether you need multiple languages. It then scaffolds a minimal Next.js project that renders each screenshot as an advertisement designed to sell the app rather than just show its interface. The whole generator ends up as a single page file that you can run in a browser, click any screenshot, and export it as a PNG at the exact pixel dimensions required by each store.

The supported export sizes cover the full range of Apple screen sizes (from 6.1-inch to 6.9-inch) and Google Play formats including phone portrait, 7-inch tablet, 10-inch tablet in both orientations, and the 1024 by 500 feature graphic. Android device frames are drawn with CSS alone, so no additional image files are needed for the Android mockups.

The skill also handles more advanced cases. You can generate screenshot sets for multiple languages, with each locale getting its own folder. Theme presets let you define two or three visual directions (for example a clean light version and a dark bold version) and swap between them quickly without rebuilding the slides from scratch. For languages like Arabic or Hebrew that read right to left, the skill instructs the agent to set RTL layout on the canvas and rewrite the copy and composition to feel native rather than just mirrored.

Installing the skill takes one command via npx, and it works across more than 40 AI coding agents. A working example app approved by Apple is linked from the README.