Selenium’s Successor, Quality vs Speed, AI in IDEs and more TGNS165

[00:00:00] Joe Colantonio Is the next generation Selenium era about to begin? What new AI agent can write and execute your test? And are you sacrificing quality for speed? Find out in this episode of The Test Guild News Show for the week of August 10th, to grab your favorite cup of coffee or tea and let's do this.

[00:00:16] Hey, before we get into the news, I want to thank this week's sponsor Zaptest AI, an AI driven platform that can help you supercharge your automation efforts. It's really cool because their intelligent co-pilot generates optimized code snippets while their planned studio can help you effortlessly streamline your test case management. And what's even better is you can experience the power of AI in action with their risk-free six-month proof of concept featuring a dedicated ZAP expert at no upfront cost. Unlock unparallel efficiency and ROI in your testing process. Don't wait. Schedule your demo now and see how it can help you improve your test automation efforts using the link down below.

[00:00:55] Joe Colantonio All right, so let's dive right into what I think is the biggest news for testers. Jason Huggins, the creator of Selenium just announced what he's calling the next generation of browser automation. What is it? Let's check it out. So on this LinkedIn post, Jason talks about how he's about to introduce Vibium, which is an AI native browser automation tool positioned as a successor to Selenium. And he also posted a video that he tied to a potential Y Combinator application. Jason recounted his experience on the tech search team that repaired healthcare.gov in 2013 including a public incident where a Keysight button failed during a congressional airing. He said that that moment highlighted the limitation of existing web testing tools at the time. With current advances in AI, though, Jason believes it's now possible to build the tool he wished existed back then. Enter Vibium, which allows tests to be ran in plain English, featuring self-healing capabilities, and aimed to eliminate brittle locators and flaky tests. And in the chat with him I had on LinkedIn, he also said the initial focus will be on JavaScript TypeScript positioning Vibium to compete directly with Playwright. So any time Jason speaks, I listen. You should definitely check out more about this in the link down below.

[00:02:10] Joe Colantonio Are you speaking of AI assisted testing tools? Imagine having an AI agent that doesn't just help you write functional tests, but actually executes them in your IDE. In this LinkedIn post, Jason Arbon told me that coTest Pilot has launched early access invites for AI test agents designed specifically for Cursor, Windsurf, and Visual Studio Code, Copilot IDEs. And this is built by the team behind Testers AI. This tool positions itself as an AI powered test team you can use inside your IDE. You highlight code, run checks with a single click and receive nearly instant feedback from several AI testers in parallel. No dashboards or stand ups required. And he also told me that the AI agents are purposely built to catch the kinds of issues commonly introduced by AI generated code, they scan for a wide variety of problems, including privacy, and security flaws like hard coded secrets or missing encryption, accessibility lapses and screen reader issues, performance concerns like memory leaks and slow queries and other critical areas like edge case handling, input validation, error logging, and production artifacts. And the really important point he told me is that the platform introduces what's called an AI engineering loop, where AI systems test AI ran code while humans focus on the remaining task and remain in the loop.

[00:03:34] Joe Colantonio If you're using Playwright for your functional test, here's an update you probably want to hear. And this came on my LinkedIn feed as well via Raj, who just announced on LinkedIn Playwright MCP YAML test has released version 2, bringing several significant updates aimed at expanding testing capabilities. This new version now supports both UI automation and comprehensive API testing in a single framework. It integrates more deeply with Playwright MCP and Gemini CLI, enabling AI assistant execution. Testers will now also see richer HTML reports and the framework can now auto-generate tests in JavaScript, TypeScript, Python, C#, and Java. And to take a deeper dive on this update, definitely check them out in the links down below.

[00:04:15] Joe Colantonio And speaking of enhancements, Report Portal just got a major AI upgrade that could change how you analyze your test outcomes. Check it out. I think I saw this on Demetri's LinkedIn page for this announcement, and it's all about how Report Portal has launched its MCP server, a middle web bridge that connects Report Portal with AI assistants like Claude, GitHub Copilot, and Cursor. The MCP Server implements the model context protocol, enabling language models to query report portal data, such as test launches, failures, logs, attachments, auto analysis, error detection, and quality gates, and as shared in the recent LinkedIn posts, QA teams and developers can now build intelligent agents with interfaces directly with the report portal ecosystem. And it also says these agents can now also perform structured AI driven test analysis and root cause detection, replacing manual log execution. Also, if you don't know, I talked about my test tool match a few weeks ago. If a tool has an MCP server, my tool will flag it and have a link to it directly. I just updated it for report portal. So if you look for a test management system. And then click on all the tools and go to report portal, you'll see it has this special MCP server button. And if you click on it, it'll take you directly to more information about how to set up the MCP server. Alright. Just a little Easter egg for everyone that's listening, that's followed along this far.

[00:05:37] Joe Colantonio All right. And I'm always on the lookout for real world case studies. And here's one all about how AI boosts performance engineering productivity. This next article by Salesforce Engineering published a Q&A with director of software engineering, Zach Zimmerman. Explaining how his data cloud performance engineering team uses AI to analyze large scale performance data, speed incident response, and build automation. And Zach states the team serves as the last line of defense for performance scalability and reliability across data cloud, supporting internal feature teams and developers before changes reach customers. And to pressure test at scale, the group runs mostly blast exercises that push data cloud to its limits and controlled environments. But AI now helps create Jupyter notebooks and visualization to inspect data to surface patterns and suggest next step actions from millions of events and by implementing the model context protocol with Cursor engineers querying internal telemetry and metadata and natural language instead of manually hunting tables and building dashboards, cutting analysis by up to three to four X, and during live instances that span different query languages, AI builds complex context aware queries, Zach reports up to 90% faster query building and engineers who are 2x faster or more during incidents. And it goes on to talk about how AI has accelerated their automation work, writing analysis pieces, APIs, notebooks and supports daily regression monitoring to keep a constant performance heartbeat on their data cloud. Now, for a reality check ever tester needs to hear, there's a dangerous myth going around about shipping fast versus shipping quality.

[00:07:15] Joe Colantonio And this is yet another awesome post by Brijesh DEB, who warns about the mantra, if you have no bugs in production, you're going too slow. And he breaks down while speed and delivery is important. He argues how that prioritizing it without a foundation of quality is risky, eroding credibility, trust, and in some instances, endangering lives. This article details sectors specific risk. I see when Citi was fined 400 million in 2021 for process failures and in aerospace software shortcuts have actually contributed the tragedies like the Boeing 737 Max incident underscoring the necessity of rigorous testing and redundancy and even a high growth tech speed first thinking fused technical debt that eventually slows innovation as Facebook discovered when it shifted from move fast and break things to move fast with stable infrastructure. And it concludes by calling for balancing speed with quality through context aware risk modules, sustainable delivery pipelines, and shared quality ownership across teams.

[00:08:16] Joe Colantonio All right. And while talking about quality, here's another thing that affects the code you're testing. Claude Code just launched automated security reviews that catch issues before you even see them. How? Check it out. All right, in this post, Anthropic has introduced automated security reviews inside Claude code. The update adds a security review command that you can run from the terminal before committing code. It basically scans your code base and explains issues it finds using a security focused prompt that checks for SQL injection risk, cross-site scripting, authentication, and authorization flaws, and secure data handling and dependency vulnerabilities. You can then ask Claude code to implement fixes. The release also includes GitHub actions that triggers automatically. On new pull requests, reviews diffs from security concerns, and post inline comments with recommended remediation.

[00:09:07] Joe Colantonio And here's another security tool that's worth knowing about. It was actually reported by my friend, Eran Kinsbruner, that talks about a new release of Checkmarx that announces Checkmarx And this tool embeds directly within AI native integrated development environment. You've probably been seeing a trend through all these new shows. And it also emphasizes that as AI begins to generate a growing share of enterprise code, traditional security methods are falling short. Developer assist places controls on the point of creation, whether the code originated from humans or AI.

[00:09:40] All right, for links of everything value we covered in this news episode, head on over to the links in the comment down below. So 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.