Welcome to the 311th issue!

Or so goes the popular saying for today.

And we've been given more force than we ever had — thanks to AI.

But with great power comes great responsibility.

And that's what I want to remind you with these three fantastic articles from tech leaders:

• Vibe Coding Works Until It Doesn't. AI-Generated PRs Carry 1.7× More Issues by Alex Svystun (CTO)
• Here Come the Bugs by Don Neufeld (CEO)
• The AI Quality Paradox by David Stockton (VP Eng)

Happy testing!

Dawid Dylowicz

Joshua Bihun introduces a score-based bug triage approach that takes both severity and priority into account.

In relation to that, Okeke Kenneth gives several handy tips on How to write a Bug Report that Developers actually act on.

What does it feel like to test alongside AI through a full QA cycle? Betty Lin kicks off a hands-on series with honest remarks on where AI helps and where instinct takes over. You can already read part 2, part 3 and part 4.

Similarly, Jason Arbon started a series: Day 1 • Evaluating How Well AI Can Find Bugs.

And testing is not quality. Michael Bolton points out a problem with calling testers quality engineers and pushes back on the rebranding trend.

Moreover, Callum Akehurst-Ryan suggests that You're not ready for Quality Engineering while Luke Lattimer-Rogers says that You don't want Quality Engineers.

Don't worry — Ujjwal Kumar Singh shares that the strengths of testers are unlikely to be replaced by AI.

However, you may want to learn from Julia Kocbek about How to approach AI and stay sane.

There are some good points brought up by Simon Prior on how to test and not to test LLM-powered applications, followed by advice: Don't Just Test the Output… Challenge the Reasoning.

There's also a great reminder from James Bach that Responsibility is the Human Moat.

Rohit Kshirsagar describes building an AI-powered solution of three agents for ticket reading, component analysis and scaffold test implementation.

Furthermore, Andrii Lysenko advises how to handle cases when Documentation exists. The agent doesn't read it.

If a test passes on retry, don't delete it. It might be telling you about a flaky bug, not a flaky test. Kevin Roe makes a strong case for retries with receipts.

Similarly, Gil Zilberfeld describes The Flaky Test Files: The Case of the State Pollution.

A handy pattern from Higor Mesquita for anyone using Playwright or Cypress, advising to move locators, waits and flow details into the model so specs are easier to understand.

Also, Swati Seela demonstrates that You Don't Need AI to Repair Your Broken Locators. You Can Do It Too.

Fernanda Nadhiftya revisits the Fast, Isolated, Repeatable, Self-validating and Timely test principles and demonstrates them with examples.

Moreover, Mochamad Syahrial Alzaidan briefly describes The AAA Pattern: A Simple Way to Write Better Tests.

Is your CI test suite getting slower? Dennis Martinez explains why this happens, which metrics to track and how to fix the worst offenders before execution time increases.

Alan Richardson shares a first look at API Spector — an open source tool for HTTP and WebSocket testing, with imports from Postman, Insomnia, Bruno and OpenAPI, plus mock servers and contract testing.

Moreover, John Ringler showcases how to perform Automated API Testing with Jest, Supertest, and Docker.

If your Playwright tests fail in CI with no request or response data to debug, Viatsheslav Pashanin suggests adding a logging middleware that attaches every API call to the HTML report.

Also, Vitaliy Potapov wrote another insightful article about Playwright in Pictures: Why Workers Restart?.

Want to try MCP servers in your testing workflow? Michał Ślęzak shows four practical use cases of popular tools and points out the security and token costs you should watch for.

Additionally, Andrew Knight shares BDD Gherkin Guidelines for AI Coding and Testing which you can find directly here.

Testing mobile apps? Ewan Mak shares a hands-on look at Maestro MCP, including an honest opinion on the 70-80% first-pass accuracy and iOS WebView gaps.

Furthermore, Prasenjit Sinha gives advice on building A Layered Mobile Testing Strategy for Long-Term Stability.

Mohammad Faisal Khatri shows how to use Playwright MCP to test the feel of an app, not just the functions. Good walkthrough with prompts, generated code and a refactoring checklist.

Similarly, Rakesh Karkare describes in detail How I Built a Local AI Browser Agent Using Python + Ollama (LLM Model) + Playwright MCP.

Alex Khvastovich continues an insightful series analysing the QA job market and shares data-driven insights from the last month in 12 minutes.

There are some great insights into testing and quality from Alan Page who recently did three podcasts across various channels.

This is AGI... 😂

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