microsoft/playwright

Playwright is a framework for automating web browsers and testing websites. The core idea is that you write code which controls a real browser, opening pages, clicking buttons, filling forms, taking screenshots, and you can do this against three different browser engines through one shared API: Chromium, Firefox, and WebKit. Whether you are writing automated tests, building scripts that scrape data, or letting AI agents drive a browser on your behalf, the same set of commands works. The way it is delivered is through several entry points that share the same engine. Playwright Test is a full test runner with built-in features like auto-waiting, where commands wait for elements to actually be ready before acting on them, and tracing, which records every step of a test run for later inspection. Playwright Library is the same browser control without the test runner, useful when you just want a script that visits a page and takes a screenshot or generates a PDF. There is also a CLI for coding agents, an MCP server which lets AI assistants drive a browser through the Model Context Protocol, and a VS Code extension. You would use Playwright when you want end-to-end tests for a web application across multiple browsers, when you need automation scripts to control a browser programmatically, or when you want to give an AI agent the ability to browse and interact with web pages. Tests run in isolated browser contexts in parallel by default, with locators that find elements the way users see them, by role, label, placeholder text, or test ID. The project is published in TypeScript and installed through npm. The full README is longer than what was provided.