本课程教你利用Playwright MCP协议和AI代理，革新2026年自动化测试工作流。从探索到CI，打造企业级端到端测试套件，捕获真实缺陷，构建四大核心AI技能。
原始标题：Vibe testing in 2026: Automated Testing with Playwright & AI
本课程的核心突破在于重塑了2026年的自动化测试工作流：它彻底告别了传统人工编写脆弱选择器和异步等待的繁琐模式，转而采用以人类为“总导演”、AI代理为“执行者”的全新架构。
学员将利用Playwright MCP协议驱动真实浏览器对含有15个刻意引入缺陷（如购物车算力错误、明文密码等）的TechShop应用进行深度探索与漏洞挖掘，并在全流程中构建用于测试设计、代码编写、缺陷报告及偶发性失败排查的四大核心AI技能，最终打造出一套能够深度融入GitHub Actions持续集成流水线、具备自动追踪与分析能力的企业级端到端测试套件。
Published 7/2026
MP4 | Video: h264, 1920×1080 | Audio: AAC, 44.1 KHz, 2 Ch
Language: English + subtitle | Duration: 2h 15m | Size: 1.88 GB
Build a real Playwright E2E suite with AI — an agent drives the browser, you direct. From exploration to CI.
What you’ll learn
Build a complete Playwright end-to-end test suite that catches 15 real bugs in a live application
Use Playwright MCP to let an AI agent drive a real browser and explore an app before any test exists
Build four reusable AI Skills that encode your standards — for test design, authoring, bug reporting, and flake triage
Read AI-generated tests critically and catch fragile locators, weak assertions, and false greens
Wire your suite into a GitHub Actions pipeline that runs headless on every push, with reports and traces archived
Triage a failure as a real bug or a flaky test, then write a bug report a developer can act on
Requirements
Basic understanding of how a web app behaves — pages, forms, buttons (no coding experience required)
A computer running Mac, Windows, or Linux
A free GitHub account
Antigravity IDE and Node.js — both free, installed in Section 3
Description
This course contains the use of artificial intelligence.
It is the whole point. This course teaches an AI-first testing workflow, and the course itself was produced with it: AI-assisted scripts, narration, translations, and code, all reviewed and directed by a human instructor. What you see on screen is what the workflow actually produces.
Playwright is the modern standard for browser testing, and end-to-end automation is one of the most requested skills in QA job postings. In this course you learn to do it the way it actually works in 2026 — by letting an AI agent drive a real browser, explore the app, and write the tests, while you stay the reviewer who decides what matters.
What makes this course different
Most Playwright courses teach you the API. This one teaches you a way of working.
Before you write a single test, you watch an AI agent open a real browser through Playwright MCP and explore the app on its own — clicking, reading, and reporting what it finds, like a tester sitting at the keyboard.
Then you build something most courses never give you: reusable Skills for the agent. Four of them — for designing test cases, writing tests, reporting bugs, and telling a real bug from a flaky one. You build them once and keep them. They are the leverage you carry to your next project and your next job.
What you will produce
– A complete Playwright test suite (TypeScript) covering login, cart, and checkout
– Your own AI Skills pack — four reusable skill files you keep forever
– Structured bug reports generated from real test failures, ready to file
– A GitHub Actions pipeline running the suite headless on every push, with HTML reports and traces archived
– A capstone project on a second app you have never seen — completed independently
– A prompt library ready to drop into your next project
Tools you will use
– Playwright (TypeScript)
– Antigravity IDE + Playwright MCP
– GitHub + GitHub Actions
– Node.js
Every tool has a free tier. You do not need to spend anything beyond the course. However if you want to avoid waiting times for token recharge I recommend to purchase Google AI package for Antigravity for one month. In some countries this first month is free in some countries it is affordable price.
The honest part about AI
The AI writes the code. It does not know what “correct” means in your application — you do. So a large part of this course is learning to read what the agent produces and catch it when it is wrong: fragile locators, weak assertions, and the most dangerous thing in automation — a test that passes for the wrong reason. Green does not mean correct. You mean correct.
Who this course is for
Manual and QA testers ready to move into automation without a coding background
Students who completed the Manual Vibe Testing course and want to automate what they tested by hand
Developers who want durable end-to-end test coverage on their web applications
QA engineers who want to see how AI agents change the speed of writing a test suite
Anyone who wants a real portfolio project — a suite, a skills pack, and a live CI pipeline on GitHub
百度网盘下载：
