Octomind

The newest << State of JavaScript >> is out. In testing, it confirmed the strong upward trend for Playwright in both adoption (usage, retention) and sentiment (interest, positivity). The respondents also used the opportunity to address their grievances with end-to-end testing. 😡    “There should be simpler ways to generate e2e tests”  “E2E tools are very heavy and slow” “I haven't found a viable way to test my front end app e2e” “Difficult setup, particularly of e2e tests” “E2E tests are painful to write” Flakiness and speed concerns had its own sub-genre. We hear you. We work diligently on making life easier for all of you. Check out the survey https://lnkd.in/dRcjt3V2 #stateofjs #javascript #testing

We build Octomind in public. An important part of it is the product update Daniel Roedler is sending bi-weekly to our users. He managed to squeeze in one before July the Fourth, last week. Tl;dr: The Good ** AI agent benchmark automation ** AI agent handles shadow DOMs & iFrames now ** We fill more robust variable templates into generated tests ** We revamped the UI for less clutter while keeping its flexibility 👇 - new layout of homepage overview and individual test cases - full size side panel - ditched tabs in favor of a single toggle button 🔛 📴 The Complicated  It dawned on us how complicated snapshots become when editing test steps. And we bumped into hydration and Azure issues. Read more 👉  https://lnkd.in/eSMX_jpH #buildinpublic #AItesting

What an honour On Deck to be on the Top Companies list. Thank you 🙌 #TopCompanies2024 #ODF18

*** Introducing a new 🐙 to our consortium ***  When your team moves fast you’ll need a good coach so everyone runs in the same direction without tripping over one's tentacles. Maike Burkard started in May as Senior Agile Coach to help us with that. “I’m here to ask the right questions. I’m here to apply the helpful parts of the agile toolbox to assist our team with overcoming challenges. I’ll make sure we are equipped with everything we need to find our product-market fit.” Welcome on board, Maike  🛟 🐙 #octoneers

Editing your tests should be a breeze. Visual locator picker just got a much needed upgrade. #buildinpublic #testing

We wrote a blog. tl;dr we shared our experience with the LangChain tool and wrote why we stopped using it. Took off on Hacker News and went on to be shared all over Reddit. The discussion was, at large, constructive. Thanks to everyone for contributing. Yet, the shebang created had an aftertaste that can be best described by one redditor’s summary “I think langchain gets way more hate than it deserves.” Internet’s gonna internet. Thank you Harrison Chase for your response. Fabian and the team really appreciated it.

No octopi got hurt in this photoshoot.

Stack Overflow asked 37K engineers about the use of AI in their development workflows. No surprise there - the trends are in line with what our users tell us. While the current use is dominated by the `writing code` use case, devs are looking into using AI for `testing code` in the future. They are curious about whether they can offload some of the testing grind to AI. Octomind is here to help with that. Btw, Stack Overflow is conducting the 2024 survey right now. Wonder how these numbers will compare YoY. #AItools #developers

Do you need a framework for building AI applications? We came to the conclusion that we did not. We use AI agents with multiple LLMs to automatically create and fix end-to-end tests in Playwright. Until just a few months ago, we did it with the LangChain framework. Fabian Both shares our struggles with LangChain and why replacing its rigid high-level abstractions with modular building blocks simplified our code base and made our team happier and more productive. Learn more 👉 https://lnkd.in/eiTc4h8P What is your experience with AI frameworks? #aiagents #aidevelopment #llm

Seems like we made a good call with building Octomind on top of Playwright. We are big Playwright believers, but we’ve built our tool in a way that makes it possible to run tests in parallel with other frameworks and expand the framework coverage in the future. We're keeping an eye on this.