web-infra-dev

web-infra-dev/midscene-skills

8 resources in this repository

GitHub
🎯8

🎯Skills8

🎯desktop-computer-automation🎯Skill

Vision-driven cross-platform UI automation skills built on Midscene.js. Supports browser, Chrome Bridge, desktop (macOS/Windows/Linux), Android (ADB), and iOS (WebDriverAgent) automation using natural-language commands and screenshot-based recognition.

desktop-computer-automation
🎯browser-automation🎯Skill

Vision-driven cross-platform UI automation skills powered by Midscene.js, supporting browser, desktop, Android, and iOS through screenshot-based control without DOM or accessibility labels.

browser-automation
🎯android-device-automation🎯Skill

Vision-driven cross-platform UI automation skills powered by Midscene.js, supporting browser, desktop, Android, and iOS through screenshot-based control without DOM or accessibility labels.

android-device-automation
🎯ios-device-automation🎯Skill

Vision-driven cross-platform UI automation skills powered by Midscene.js, supporting browser, desktop, Android, and iOS through screenshot-based control without DOM or accessibility labels.

ios-device-automation
🎯harmonyos-device-automation🎯Skill

A vision-driven cross-platform UI automation skill built on Midscene.js that uses natural-language commands and screenshots to control browsers, desktops, Android, iOS, and HarmonyOS devices, with support for E2E test scaffolding.

harmonyos-device-automation
🎯chrome-bridge-automation🎯Skill

Vision-driven cross-platform UI automation skills powered by Midscene.js, supporting browser, desktop, Android, and iOS through screenshot-based control without DOM or accessibility labels.

chrome-bridge-automation
🎯vitest-midscene-e2e🎯Skill

Vision-driven cross-platform UI automation skills built on Midscene.js. Supports natural-language driven control for browser, desktop (macOS/Windows/Linux), Android, iOS, and HarmonyOS platforms, operating entirely from screenshots for reliable cross-platform automation.

vitest-midscene-e2e
🎯midscene browser automation🎯Skill

Skill

midscene browser automation