Workshop: Advanced Web Testing with Playwright and AI

on YouTube ↗01h 31m

" summary

A build-along that actually shows the workflow: Playwright + local CLI skills (not MCP) set up from scratch, Chromium-only config, and a live run against the Buggy Board app in Cursor that produces plan, generate, heal test plans from specs, then generates 13 login tests implemented with page objects and fixtures, runs them locally, and wires CI advice (parallel workers, caching, atomic tests). Details include CLI skill files, seed auth spec, and when to review AI healing.

AI-written summary · about

" chapters (35)

1. 0:00Workshop Opens - Advanced Web Testing With Playwright and AI
2. 2:08Speaker Intro - Andrew Knight, Senior Director of Product & Engineering at Cycle Labs
3. 5:43What Playwright Is and Why the Debug Protocol Makes It Fast
4. 6:44Browser Projects vs Big-Box Browsers, and Why That Matters
5. 7:45Built-In Goodies: Auto-Waiting, Codegen, UI Mode, API Testing
6. 8:46Playwright Is Natively AI Supportive - MCP, Agents and a CLI
7. 13:23Selenium vs Cypress vs Playwright, Fairly Compared
8. 16:25The AI Stack Defined: Coding Agent, Model, Rules, Sub-Agents, Skills
9. 19:37MCP vs CLI - and When to Choose Each
10. 22:11The Three-Step AI Testing Process: Plan, Generate, Heal
11. 24:12A Warning on Healing: Review Failures Before You Let AI "Fix" Them
12. 26:45From Prompt Engineering to Context Engineering
13. 29:19Spec-Driven Development: If You Can Describe It, You Can Do It
14. 33:58Turning Best Practices Into Rules: Page Objects, Fixtures, Test IDs
15. 37:36The Kingpin Rule: Generate Tests Whenever You Generate Product Code
16. 40:07Three-Way Programming: Developer, Tester and AI
17. 42:10The Test Automation System: CI Is the Production System for Automation
18. 44:46The Five-Step Playbook for Scaling a Test Suite
19. 46:50Live Demo Begins: The Buggy Board App in Cursor
20. 51:30Asking AI to Explain an Unfamiliar Codebase
21. 53:33Inside the Specs Folder: The Project Constitution
22. 57:35Installing Playwright Into the Project From Scratch
23. 1:00:40Configuring Chromium-Only - Manually or by Prompt
24. 1:03:14Installing the Playwright CLI and Skills
25. 1:06:53Invoking the Playwright CLI Skill Directly
26. 1:07:24Prompting for Test Plans From Specs Plus Live App Exploration
27. 1:11:37Q&A: Prioritising Tests From Recent Code Changes
28. 1:12:08Q&A: Is the Generated Test Checking Business Behaviour or Just the UI?
29. 1:14:45Reviewing the Generated Test Plans Per Feature
30. 1:15:47Where AI Finds Negative Cases Humans Skip
31. 1:16:18Generating the Login Tests From the Plan
32. 1:19:01Q&A: Where AI Genuinely Improves Playwright Beyond Boilerplate
33. 1:21:37Q&A: Determinism for Enterprise CI/CD - and Humans Aren't Deterministic Either
34. 1:23:39The Result: 13 Login Tests With Page Objects and Fixtures
35. 1:26:11Q&A: Biggest Challenges at Scale; Migrating From WebdriverIO

" related talks

• 48mSupercharged Testing: AI-Powered Workflows with Playwright + MCPDebbie O'Brien · NDC London 2026
• 35mState of end-to-end testing for modern web applicationsVlad Filippov · REFACTOR DX 2023
• 01h 31mAdvanced Playwright with AIAndrew Knight · TestMu Conference 2025
• 39mHow to Use AI with Playwright for UI & API TestingMichal Slezak · TestCon Europe 2025
• 14mAdvanced Playwright Techniques for flawless testingDebbie O'Brien · TestFlix 2024
• 25mAI-Powered Debugging & Browser Automation with Playwright MCPDebbie O'Brien · Frontend Nation 2025
• 41mTesting Web Applications with PlaywrightDebbie O'Brien · Build Stuff 2022
• 13mHow to Use Playwright for Automating Accessibility TestingSundaresan Krishnaswami · TestFlix 2023