Welcome to the 310th issue!

The most recent ThoughtWork's Technology Radar is out and it's worth checking out.

It's a biannual snapshot of recommendations for tools, techniques, platforms, languages and frameworks from one of the most respected consulting companies.

It's great not only for spotting trends but also for identifying potential risks and dangers.

And if you're wondering what it means, Patrick Prill wrote a great summary — The Tech Radar is Blinking Red.

Happy testing!

Dawid Dylowicz

How does a quality engineer's job actually change with experience? Shubham Sharma looks back and reflects on the perception of quality, imposter syndrome and why senior engineers tolerate uncertainty better.

Moreover, Swati Seela shares her point of view on What Makes a Good QA Lead.

The author, whom I unfortunately couldn't identify, shared a really good, in-depth article with practical advice on how to think about each layer of tests the right way.

Similarly, Keith Klain brings up a few good points on Thinking, Fast and No.

Are your AI-generated tests actually catching bugs? Simon Prior prompts quality leaders to use AI as a thinking partner for risk and strategy rather than a faster way to write flaky tests.

Simon also wrote a good article about Quality Culture: Shared Ownership Is Harder Now, Not Easier.

A great, thoughtful read from Vernon Richards on how the tester role is evolving from hands-on work to coaching the team and guiding AI agents.

At the same time, Gary Hawkes explains Why Quality Engineers Matter in the AI Era.

Patrick Prill observed a funny pattern — Spec-Driven Development asks devs to write perfect specs upfront, which is the same thing we admitted we were bad at during the Waterfall era, before Agile came.

So where does it lead? Simon followed up with another article: Are We Building the Right Thing, or How GenAI Made Us Forget to Ask.

If you're testing AI-powered solutions, Jeff Nyman has a practical guide on using DeepEvals to check if a model holds its thread across a full multi-turn conversation.

Want to stop AI from writing shallow API tests that just check for a 200 response? Wendy Erdheim-Poch shares a step-by-step workflow that scans routes, asks the right questions and reviews its own output.

Furthermore, Josphine Job shares good advice on Why Every QA Should Build Their Own AI Workspace.

While it's easy to comment out a broken test, it's never a good idea. David Mello demonstrates how to use a skip pattern instead with bug ticket links and ready-to-use snippets for Jest, Playwright, pytest and JUnit.

Moreover, Swati Seela wrote a thoughtful article about Retries in Test Automation: Safety Net or Silent Problem?

Tired of trying to figure out if a failing test is flaky or showing an actual problem? Giuseppe Donati shares how trivago built an AI helper that does that work for them.

Chris Kenst inherited a 2,500-test Cypress suite that takes 45 hours to run and explains why automating everything leads to pain.

Similarly, Jitesh Gosai explains well When quantity leads to quality.

Want testing to start on day one instead of waiting until the end? David Ingraham and Leonardo Lanni wrote a thoughtful article explaining how a UI skeleton with mocked data lets front-end, back-end, and QA move at the same time.

On that note, Marko Churlinoski explains why we should Stop Writing New Tests. Supercharge the Ones You Already Have.

Anton Gulin describes Playwright v1.59.0's new Screencast API, which records the session as a video with chapter markers and action labels that you can attach to a ticket.

Can you make AI work in a loop to automatically close the coverage gap? Chandni Kaithavalappil tried Karpathy's autoresearch on a small app and shares what happened.

Moreover, Irfan Mujagić explains how to go From CI Failure to Root Cause in Seconds: MCP for QA Engineers.

If you don't want to write login steps in every Playwright test, Aparna Mishra breaks down how fixtures pair with POM to handle the test setup for you.

Also, Can Claude Skills Replace Playwright Agents? A Practical View for QA Engineers by Kailash Pathak.

Vitali Haradkou built three helpful solutions that bring SQL validation to Playwright tests, including compile-time checks on SQL queries, parameter inference from placeholders and automatic connection teardown.

Additionally, Reyad Hassan explains How to Mock APIs and Own Your Test Scenarios with Playwright.

David Ingraham outlines why Cypress suites get gradually slower and shows a few tweaks that bring fast feedback back to your team.

Similarly, Vitaliy Potapov describes in detail how the Fully Parallel Mode in Playwright works.

If you're curious about what agentic AI testing solutions are on the market right now, Daniel Knott gives a 15-minute high-level overview.

Playwright 1.59.0 is out with some exciting features that Artem Bondar discusses in this 19-minute recording.

Similarly, Alex Khvastovich also did his own overview of the Playwright 1.59 New Features You Should Know.

If you're doing test automation, you're not really testing... 💡

Thanks for reading!

If you enjoy this newsletter and find it helpful in becoming a better tester, please consider sharing it with others.

Dawid Dylowicz