Playwright + AI, Faster Migrations, Smarter Tests and More TGNS176

[00:00:00] Joe Colantonio What tool is trying to give testers more control over flakiness and element changes without rewriting everything? How do you migrate Cypress to Playwright? Find out on this episode of the Test Guild News show for the week of November 30th. Grab your favorite cup of coffee or tea, and let's do this. First up is a new tool I've been hearing more about lately, and that is Test Driver. They have Test Driver Studio, which is tackling one of testing's most persistent challenges, the time and complexity of writing Playwright tests from scratch. This platform now enables teams to generate and execute tests directly from plain language descriptions, eliminating the traditional scripting bottleneck a lot of us face. Test Driver Studio automates the entire test creation workflow. Teams simply describe what they want to test in plain English, and the platform generates the corresponding Playwright test instantly. The system then executes these tests in real browsers and can schedule them for continuous integration and delivery pipelines, all without manual code. This platform addresses a specific pain point, maintaining test scripts when UI elements change. Instead of updating selectors and rewriting test code, users simply update their plain language descriptions, and test driver regenerates the test automatically. This approach reduces test maintenance from hours to possibly minutes. As you can see, Test Driver offers full access to demonstrate the platform's capabilities. The system supports both UI flow validations and end-to-end test scenarios, executing tests across real browser environments. Rather than simulated ones. And for teams already invested in Playwright but struggling with test creation speed or maintenance overhead, Test Driver Studio positions itself as a bridge between manual test writing and full automated test generation. The platform maintains Playwright testing power while removing the coding complexity. Sounds awesome, right? Try it for yourself for free using that special link down below. Let me know how it goes for you.

[00:01:55] Joe Colantonio All right, a lot of times I hear people want to know how to migrate from Selenium to Playwright. But how about Cypress? If that's you, here's a new resource that helps you do this. Karthik just posted a new video that shows a real Cypress test being executed, then passed into an AI agent, which converts the entire script, including locators, testing logic, and file structure, into Playwright. In the video, Karthik then runs the newly generated Playwright test, which executes successfully and produces a complete test report. In the video, he also demonstrates covering converting tests into a BDD format, and his video concludes with confirmation that the same method can be used in more complex frameworks, as long as testers give the AI enough context about custom code they need to be preserved during migration. So I don't know if you're like me, you've probably been hearing a lot about goose. Here's an article how to use Goose with Playwright MCP. And this posted by Filip who documents how he automated a ticket submission task using Goose, which is an open source AI agent from Block combined with Playwright's MCP in a Chrome bridge extension. And according to the article, the goal was to solve the common problem of converting handwritten notes and session feedback into tickets after focused work periods. Goose is described as an AI assistant capable of executing computer actions, not just generating text. It works through extensions such as MCPs, browser tools, and other integrations that allows it to browse the web, run code, and interact with APIs. Goose Desktop app lets users choose their preferred AI model. The article explains that the Playwright MCP extension enables Goose to automate browser actions like opening pages, clicking elements, and filling forms out. It operates in a continuous feedback loop that allows it to find interface elements even when layouts change without the strict selectors required into digital automation. Filip also highlights authentication as a key challenge since Playwright MCP normally launches fresh browser sessions that require login each time. The Playwright MCP Bridge Chrome extension solves this by connecting Goose to an already logged in browser supporting SSO and 2 factor authentication. He also goes over the setup of using the Chrome browser extension. Once configured, Goose could follow verbal instructions using tools like Super Whisper to add issues to a tracking system while the automation runs in the background. Really cool demonstration. If you haven't followed Filip, I highly recommend you should follow him and check out more from him using the link once again down below.

[00:04:24] Joe Colantonio Maybe saying enough about Playwright, I have mobile apps that need to be tested. What can I do with AI? Well, here's one article you should definitely check out. It's a post published by Vince, who described a new AI assistant mobile testing capabilities that's been added to the WebDriverio MCP server. And according to him, the update extends the existing browser automation support to mobile applications by adding Appium based automation for both iOS and Android. On LinkedIn, Vince also explains that the same low token approach used for browser testing now applies to mobile workflows, allowing testers to launch apps, handle permission dialogs, and search within apps through natural language prompts. Vince also says this works connecting WebDriveio's cross-platform automation with LLM-driven testing, and can be combined with sub-agents to build test suites. He also says that this capability functions on emulators, simulators, and real devices, and that the goal is to make AI assistant mobile testing practical without heavy token usage. And he also pointed to the documentation for the MCP server, which includes everything you need to know, includes starting and closing sessions, navigating pages, retrieving visible elements, interacting with elements, taking screenshots, handling cookies, performing mobile gestures, all the things you need to know, as well as installation steps as well.

[00:05:40] Joe Colantonio As you know, I love automation, not just functional automation, but automation for all the things, including security and performance testing. Here's one article that talks about a new tool I just heard about called OstorLab, a new tool I just learned about that introduced a new AI penetration testing engine built specifically for mobile applications, aiming to overcome the limits of traditional manual security testing. And this article talks about how the engine learns how apps behave, manages authentication and session constraints on its own, and delivers validated vulnerability findings complete with proof of concept evidence. For teams that can typically test only a small portion of the mobile apps each year, this approach opens the door to continuous assessments across their entire portfolio. When high-risk issues surface, whether zero days or production misconfigurations, this engine can deliver actual results in hours instead of weeks that manual assessments usually require. And one of the standout features it calls is the reduction of false positives. The engine validates issues with screenshots that shows exactly how a vulnerability can be exploited. And in the article, they quote an engineering manager that notes that once developers can see the exact attack path, they stop pushing back on security fixes. And instead of monthly reports, this system generates prioritized tickets that plug directly into existing ticketing tools.

[00:06:58] Joe Colantonio All right, another article that caught my attention was on LinkedIn. This is by Benjamin. And in this LinkedIn article by Benjamin, he covers how testers who understand software architecture are able to find defects that requirement focused testing often misses. And it goes on to compare two types of testers those who simply verify requirements and those who seek to understand how a system works beneath the surface. Benjamin says the latter group uncovers deeper issues because they see how application layers interact. He explains software architecture as the structure of an application across the front end, backend, data layer, and integration layers. And according to the article, this knowledge helps testers avoid evaluating features in isolation. Benjamin also points to scenarios where a feature appears to work in the UI, but fails in the database, cache, or connected services. The article states that architecture awareness expands test coverage by revealing dependencies not outlined in requirements. He also notes that many defects occur at integration points, and testers who understand how systems communicate can identify problems earlier. He also describes how architecture diagrams guide exploratory testing by exposing redundancy risk. He also advises testers to ask a developer for an architecture view of a feature they're testing and create a new test case based on that diagram.

[00:08:16] Joe Colantonio Last up is the webinar of the week. In our Test Guild Webinar of the week, the folks on Blink.io will be joining us again to demonstrate how AI can generate functional tests directly inside an active Playwright repository. And as you know, traditional automation still depends heavily on human-created scripts, which can increase test debt and make coverage difficult to maintain. This webinar positions AI-driven test generation as a way to close the gap between test intent and implementation. The webinar will show AI producing Playwright tests in real time, including creating reusable locators, maintainable code based on natural language instructions. Attendees will see how AI interprets application flow, builds tests, validates them. And commits them into a project. It also describes how vibe testing keeps test tweets aligned with rapid product changes through autonomous maintenance and self-healing.

[00:09:08] Joe Colantonio All right, and for links of everything value we covered in this news episode, hand it over to all those links in the comment down below. That's it for this episode of the Test Guild News Show. I'm Joe, my mission is to help you succeed in creating end-to-end full stack pipeline automation awesomeness. And as always, test everything and keep the good. Cheers.

[00:09:26] Joe Colantonio Before we go, don't forget, if you want to modernize your testing, check out our sponsor, ZapTest AI's AI-driven platform that supercharges your automation efforts. Their intelligent co-pilot generates optimized code snippets, while their planned studio streamlines test case management effortlessly. Experience the power of AI in action with their crazy risk-free six-month proof of concept featuring a dedicated Zap expert at no upfront costs. Support the show and check it out for yourself by going to testguild.me/zaptestnews.