Welcome to the 288th issue!

A lot of major services were down today.

The reason? An outage of the us-east-1 AWS region.

To me, it's a good reminder to invest in Chaos Engineering to prepare for unexpected system disruptions.

For that, I recommend starting with Awesome Chaos Engineering — a great collection of resources.

Happy ~~testing~~ breaking! 🙂

Dawid Dylowicz

Building and testing AI apps requires a different approach than other types of software. Jarad DeLorenzo advises how to address the challenges with the non-deterministic behaviour.

Moreover, Keith Klain shares a sobering opinion on the ever-present Software Testing GriftopAI.

If you want to increase your salary, Mirza Šišić gives tons of tips on positioning yourself for better negotiation and interviews.

How mature is your testing approach? Toyer Mamoojee showcases the benefits of investing in the more efficient tests on the lower level and advises how to get there.

At the same time, Katja Obring has some good points on Why the Experimenter's Mindset Outlasts Automation in Software Testing.

Maria Kedemo explains why effective testing depends on understanding and reflection rather than templates, and how heuristics can help with that.

Similarly, Yanming Zhai outlines several points about QA as Systems Thinking: Applying Charlie Munger's Multidisciplinary Model to Software Quality.

If you're thinking about using AI in testing, Yerem Khalatyan points out a few use cases to explore, such as test case generation, automation and execution.

Similarly, someone asked the testing community about AI solutions for Test Case creation and they shared their setups and thoughts.

Daniel Delimata demonstrates an automated approach to test system outcomes that are only visible after some time.

This is yet another example of where AI agents can be used for testing practices. Utku Kılınçcı describes four use cases they implemented in their company.

Similarly, Aliaksandr Melnik outlines How we use Claude Agents to automate test coverage with promising results.

Do you really know what page objects are for? Đinh Công Cảnh offers an interesting thought exercise with advice.

If you worked in testing for some time, you might have noticed that it's easy to misinterpret testing-related terms, such as types of tests. Cassandra H. Leung provides more clarity on that.

Cursor is a popular AI-powered IDE, which Filip Hric tested out and shared how it works with implementing Playwright tests.

Similarly, Michał Ślęzak investigated — Cursor & Playwright: Is This the Best AI IDE for Test Automation?

Wondering what to use to test AI? Tarek Oraby created a handy list of open-source tools designed for evaluating LLMs.

Jennifer Shehane from Cypress has announced that Cypress Studio is now enabled by default and will soon gain AI-powered test suggestions.

Similarly, S Chathuranga Jayasinghe describes that Cypress Gets an AI Brain: Meet Cy.Prompt().

If you're looking to improve your Cypress test design, execution speed, or maintenance, Omer Keskinkilic has some practical tips based on their experience.

Last week, I highlighted a significant release of Playwright Agents. Sławomir Radzymiński created a great deep dive into how they actually work.

Also, Kailash Pathak did a detailed investigation to demonstrate How AI-Powered Playwright Agents Fit into the Traditional STLC.

The developer tools in browsers such as Chrome or Firefox have always been great for debugging and testing. Daniel Knott recorded an insightful, 17-minute overview of several features that you may find helpful.

Are you considering migrating to a new test automation tech stack? Alan Richardson shares some good advice on what to take into account.

First day at Amazon... 😅

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