Visual locator picker, virtual scrolling and getting to test results really fast

Hope you're doing well! I'd like to share the new batch of updates. It took a bit longer than usual, but it was worth it. We had our quarterly team get-together and decided to make room for more brainstorming rather than just crunching code.

The Good

• ‍We add dependencies automatically now. Those pesky cookie banners can be annoying for users and automated tests alike. We now add a cookie banner test as dependency automatically. More common prerequisite tests for other flows are coming soon. Looking at you, log-in! 🍪

• Mission smoother test editing is underway. You can switch interaction types, and our new visual locator picker is a treat to use. Check it out! 🍭
• Snapshots are helpful only when up to date. We’ll let you know when they are outdated. 📸
• We mark broken locators now. A yellow highlight and an indicator will point to the locator that doesn’t do the thing it’s supposed to. 🟡

‍

The Bad

We initially served automatically discovered link tests as a low hanging fruit during sign-up. Everyone needs to know their links work, right? 

We went a bit overboard with it. Fair enough, most apps are link-heavy. But soon we realized that we didn’t distinguish between static and dynamic elements. We generated too many link tests on dynamic elements that broke constantly. We decided to pause it,  rethink our approach and come back with a better link test solution. 

‍

The Complicated

Remember our struggles with shadow DOMs and nested-nested iFrames? Enter virtual scrolling. To increase the likelihood of successful test step generation we want the AI agent to handle virtual scrolling. Based on the sighs and expletives in PR comments, it’s not a trivial nut to crack.

In other complicated stuff news - we're overhauling our product analytics. Since the beginning, we've used Posthog dev tooling to streamline our UX, but we didn’t give it much love. It's time for a proper analytics setup to handle our growing feature set and user base. This will help us avoid getting lost in the data and ensure we don't lose you.

‍

The Highlight

We made a huge step in speeding up the time to your first test results and we culled the number of clicks to get there. For every successfully discovered test case we automatically generate test steps and code, validate it, and publish it to your test suite. Just hit the ‘run test report’ button.

 This is in line with our efforts to be smarter with your input for the AI loop. Your time and attention is precious, but we want you to be still in control.

‍

---

‍

Thanks a ton for your responses over the past three weeks. Many of you have taken the time to share your feedback on the product and our communications. 

We truly appreciate it and never take it for granted. Feel free to reach out to me or schedule a call directly. Whatever you have to say, we're here to listen.