Testing with Agents Workshop

Welcome to the Testing with Agents workshop! This hands-on workshop will guide you through integrating AI agents into your testing workflow using Playwright, GitHub Copilot CLI, and MCP (Model Context Protocol).

Workshop Overview

This workshop consists of 5 progressive labs that take you from basic testing with AI assistance to advanced agent orchestration. You'll work with a T‑Rex Runner application (TypeScript + Node) as your testing target.

Lab Structure

0. Prerequisites - Setup and installation requirements
1. Workshop 1: Background Agent and Playwright Basics - Set up Playwright with TypeScript and use a background agent to generate your first UI test
2. Workshop 2: Copilot CLI and Skills - Use GitHub Copilot CLI to iteratively author tests and create reusable skills
3. Workshop 3: Copilot SDK Custom Agent - Build a custom testing agent with the GitHub Copilot SDK
4. Workshop 4: MCP + Playwright Tools - Configure the Playwright MCP server to let agents browse your app deterministically
5. Workshop 5: Cloud Agent CI Orchestration - Delegate test execution to cloud agents and integrate with CI/CD

What You'll Learn

• Setting up Playwright (TypeScript) for frontend testing
• Using AI agents to generate and maintain test code
• Creating reusable skills for test automation
• Building custom agents with the GitHub Copilot SDK
• Using MCP for deterministic browser automation
• Orchestrating testing workflows with cloud agents
• Integrating agent-driven testing into CI/CD pipelines

Prerequisites

Before starting, ensure you have:

• Node.js 20+ installed
• VS Code with extensions:
  • Prettier
  • GitHub Copilot
  • GitHub Copilot Chat
  • Playwright Test (will be installed during workshop)
• GitHub Copilot CLI (will be installed in Workshop 2)
• Basic knowledge of TypeScript and testing concepts

The Application

You'll be testing a T‑Rex Runner game - a canvas-based browser game with:

• Frontend: Vite + React + TypeScript
• Backend: Express + TypeScript (REST API for high scores)
• No existing tests (you'll create them during the workshop!)

Getting Started

1. Download the workshop project (link provided by instructor)
2. Complete the prerequisites setup
3. Begin with Workshop 1

Workshop Format

Each workshop follows this structure:

• Goal: What you'll accomplish
• Story: Real-world context and motivation
• Steps: Hands-on instructions
• You learned: Key takeaways
• Stretch ideas: Optional advanced challenges

References

This workshop is inspired by the GitHub Samples "Agents in SDLC" approach and follows similar workshop cadence and guided lab style.

Support

If you encounter issues during the workshop:

1. Review the troubleshooting section in each lab
2. Ask your workshop facilitator
3. Check the official documentation links provided in each workshop

Happy testing! 🧪🤖