Welcome to the 331st issue!

Want to learn Playwright better?

You may like this: Playwright Playground.

Someone on Reddit built a free platform with 40+ Playwright exercises. There's even a lab for flaky tests.

Happy testing! 🙂

Dawid Dylowicz

When an AI agent writes both the code and the tests, you need to make sure the validation is done well. Kochi Chuang shares ways to achieve that before merging.

Moreover, Jitesh Gosai asks a similar question: When AI says it's done, how do we know?

This is a helpful guide from Stuart Thomas on how to test AI products, with advice on test strategy, using synthetic personas and leveraging the LLM-as-a-judge approach.

Also, here's a helpful read on Making Non-Deterministic AI Testable by Pankhuri Chaturvedi.

As more companies create AI assistants and chatbots, this article by Kristin Jackvony gives relevant advice on how to approach testing them.

At the same time, Lialia Sakhno has some tips on building an Evaluation suite, or "Congrats, your model is just too weak".

AI agents can write a lot of code that looks fine at first. A developer on Reddit explains how quality still gets worse over months, even with TDD, and they ask others how to fix it.

On that note, Sergii Vashchenko shares experience from Managing distributed QA teams across time zones: Lessons from 10+ projects.

For years, we turned testing into test cases, automation and dashboards. Now AI is good at all of it. John Gluck thinks about his 25-year career and asks where that leaves us.

Similarly, Josphine Job describes What Is Happening to Software QA Jobs in 2026.

Audrey Moulins describes her experience of building regression tests and gives a four-step method to generate tests you can trust.

Additionally, Michael Angelo Concepcion has a good reminder of the same problem: My AI Passed Every Test. It Was Still Wrong.

Dr. Josh Simmons describes how mutation testing can help you check the correctness of tests written by AI, with a small example.

Furthermore, another way to check AI-written tests is Red, Green, Remove: Outside-In Tests for Coding Agents, as Ian Maurer suggests.

Eleonora Belova shares the approach they took to migrate about 200 Playwright tests from C# to TypeScript, redesigning the test framework on the way. The suite now runs 30% faster with fewer flaky failures.

Speaking of that, David Ingraham reminds us that Your Test Suite Is a Product. Start Treating It Like One, while Ravi Rawal explains why Test Automation Is Software Engineering.

Michał Krassowski shows how JupyterLab made its screenshot tests work the same on a laptop and on CI, including a bot that lets any contributor update reference images with a single comment on their pull request.

In relation to that, Yanshi Tyagi has an interesting story about how I Taught Claude to Test Our App by Letting It Watch Me Use It.

If you're working with the microservices architecture, Julia Pottinger advises on how contract tests stop one service's change from breaking another before any side is deployed.

What's more, Thejaswini Arun asks a good question for API testing: You Have No Spec and a Live Production API. How Do You Know What "Correct" Is?

That's a helpful article from Sneha Rode on using Chrome DevTools for testing. For example, did you know that the Network tab can tell you whether a bug is in the frontend or the backend?

Moreover, if you test accessibility, Sebastian Clavijo Suero describes the WICK-AG-Lens Chrome Extension to Target Accessibility Issues.

Worried about flaky AI-generated tests in your suite? Vineet Verma describes a CI gate that rejects weak selectors before the tests get merged. This is followed by another article on How to Validate AI-Generated Playwright Tests Before You Merge Them.

Sonu Gupta shares how they built two AI agents for an Appium project, one to debug failing tests and one to write page objects.

Tayguara Dias Reis outlines a Playwright pipeline in GitHub Actions and points out that sharding doesn't necessarily save time when the browser install takes longer than the tests.

On top of that, Irfan Mujagić provides tips on Testing Webhooks with Playwright.

Using WebdriverIO? Christian Bromann lists out the changes in the latest version, where a coding agent can now check its own work in a browser, on a phone or in a desktop app and save the steps as a regular test.

At the same time, if you use Cypress, Jennifer Shehane shows how to Find elements the way your users do with cy.getByRole().

Mike Harris shares lessons from the Leadership is Language book, including how to pause work to make better decisions and how to ask questions that don't make people defensive.

Playwright 1.64 is out bringing some new features and Alex Khvastovich doesn't waste a minute to tell us more about it in this short video.

Similarly, Fidaa Berrjeb wrote about the same release: Playwright 1.64: The Update That Finally Shuffles the Deck

Looking for an open-source solution for performance testing? Joe Colantonio invited Abhinav Premsekhar and Lars Holmberg to talk about Locust and its capabilities which turned into an interesting, 36-minute conversation.

Heard of JDD yet?... 😂

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