Hello everyone, and welcome to Test Automation University!
My name is Rex Jones II, and I am an SDET — a software development engineer in test.
Today, we are embarking on a journey that represents a paradigm shift in the history of quality engineering. If there is one consistency across my years in IT, it is that our industry does not stand still. Right now, we are witnessing an evolution: the move away from traditional automation and into the era of agentic intelligence.
This Playwright and Agentic AI course will provide you with an engineering blueprint to scale this new wave of autonomous automation.
# Course Overview
Over five "stations," we build an agentic test framework piece by piece — from the governance rules that keep AI-generated code safe, through the Model Context Protocol servers that let a language model actually drive a browser, all the way to a token-efficient assembly line that turns an enterprise requirement into a reviewed Playwright test.
Chapter 1: The agentic shift from scripting to orchestrating, the "five-star kitchen" governance standard and the mandatory
// Created By AItag, plus the local environment setup.Chapter 2: The Model Context Protocol as a bridge — how the Playwright MCP server translates high-level intent into deterministic browser actions using the accessibility tree, and where it hits a wall.
Chapter 3: The Playwright Test MCP tri-agent lifecycle — the Planner as strategist, the Generator as SDET, and the Healer as a guarded communicator that never touches your application source.
Chapter 4: The token crisis, and why the Playwright CLI exposed as SKILLs is the production engine — planning and generating tests from documented test cases without dumping accessibility trees into the model context.
Chapter 5: The architect's perspective — the 80/20 automation split, the mechanical difference between a skill file and a sub-agent, and orchestrating a full assembly line with a human-in-the-loop validation gate.
# What You'll Learn
- Shift your role from writing every line of test code to defining goals, guardrails, and system logic.
- Enforce framework governance so AI-generated code never reaches production unreviewed.
- Configure and use the Playwright MCP server, and understand why the accessibility tree beats pixels and screenshots.
- Drive the Playwright Test MCP Planner, Generator, and Healer agents against real business requirements.
- Use the Playwright CLI as a token-efficient production engine for test planning and generation.
- Connect an enterprise quality platform (Inflectra SpiraPlan) into the agentic lifecycle via MCP.
- Design skill files and sub-agents that compose into an end-to-end agentic assembly line.
# Prerequisites
- Introduction to Playwright
- Advanced Playwright
- Familiarity with TypeScript and the Page Object Model design pattern.
