Presenters (Day 1: Workshops)

Building a Thinking Skills Toolbox for Testers

Speakers: Fiona Charles

Overview: The American baseball player Ted Williams once said, “If you don’t think too good, don’t think too much.”

Not thinking too much isn’t an option for testers. We are people who think for a living. We can’t perform effectively without thinking. Most of the other skills we need grow from our ability to think. It is our stem cell skill.

Our ability to think—and think well—is crucial when we are dealing with AI agents or the products of “vibe” coding, when we are testing AI-based systems, or when we are using AI tools to develop tests. We can’t safely outsource our thinking to technology that mimics thinking, but doesn’t actually do it.

To think well, we need to learn and practice thinking in different ways: to explore and master diverse thinking tools and techniques so we can approach problems from different directions and find novel solutions. It’s essential to hone our critical thinking, but also to become skilled at visual thinking and thinking creatively.

There are many good tools and techniques available to help us think better. We’ll explore and practice several of the most useful in this highly interactive workshop. Join us to add tools to your thinking skills toolbox, and have fun while becoming a better thinker.

Bio: Fiona Charles is a consultant and international keynote speaker specialising in the human side of software quality and testing. She designs and conducts experiential workshops at conferences, online, and in-house for clients; has recorded several webinars, and published many articles in quality-related publications. She edited “The Gift of Time” (2008), a book of essays celebrating the work of Gerald M. Weinberg. In 2019 she was awarded the EuroSTAR Testing Excellence Award.

Follow her on BlueSky @fionaccharles.bsky.social

From generation to self-healing: The full lifecycle of E2E tests with Playwright and GitHub MCP

Speakers: Dawid Pacia

Overview: A technical workshop that guides participants through the full lifecycle of E2E tests – from their generation, through CI/CD deployment, to automatic diagnostics and repair after a failure occurs. We will build a framework based on Playwright and TypeScript, optimized for cooperation with AI. Next, utilizing the free version of the Cursor environment and the GitHub MCP (Model Context Protocol), we will automate the process of analyzing logs and code. Participants will learn how a language model can diagnose, test, and pipeline errors directly based on the commit history in the developer repository, and then safely refactor the code.

Structure of the workshop:

  1. Architecture of E2E tests for AI (Playwright + TypeScript)
  • Configuring the Playwright and TypeScript environment for collaboration with LLMs.
  • Implementing the Page Object Model (POM) pattern using strong typing and JSDoc annotations as declarative prompts.
  • Generating assertions, test logic, and API mocks (page.route) in Cursor
  1. Execution in CI/CD and regression simulation
  • Configuring the workflow in GitHub Actions (.github/workflows/playwright.yml).
  • Introducing “breaking changes” into the application code (modifying the frontend DOM structure and changing the API contract) alongside a configuration error in CI.
  • Analyzing artifacts from a failed build: interpreting GitHub Actions logs, the playwright-report, and decomposing data from the Trace Viewer.
  1. GitHub MCP (Model Context Protocol) integration in the debugging process
  • Configuration and authorization of the GitHub MCP server in the local environment (Cursor IDE).
  • Building contextual queries using MCP tools (e.g., github_search_commits, github_get_pull_request).
  • Analyzing source code differences (diffs) from within the IDE: identifying removed data-testid attributes and modified endpoints directly from developer commits.
  1. Automatic refactoring and infrastructure Self-Healing
  • Updating TypeScript interfaces based on API contract changes detected by MCP.
  • Automatic refactoring of broken selectors in page objects (POM).
  • Validating AI fixes using the TypeScript compiler (tsc).
  • Fixing the YAML configuration file for GitHub Actions using context from CI errors.
  • Running the updated process and verifying its correctness (green build).

Bio: QA and Test Automation Manager as well as mentor and trainer. Tech freak following all the newest technologies (and implementing them on his own). Fan of the Agile approach to project management and products. Supporting companies in transformations toward better quality. Actively speaking and traveling around the world, sharing knowledge and experience.