Early

Big news for EarlyAI and our users! 🌟 We’re thrilled to announce another significant milestone—Vitest support is now available! This expansion means developers can generate tests with Mocks, happy paths, edge cases, and high coverage for this additional framework. To put it to the test, we ran EarlyAI on the amazing open-source project Mock Service Worker (MSW), created by Artem Zakharchenko. MSW is a seamless REST/GraphQL API mocking library for browsers and Node.js. As a result, we generated 637 unit tests: ✅ 458 passing tests ❌ 179 uncovered bugs EarlyAI is free for open-source software, and we’d love to collaborate with the OSS community. Huge thanks to our team for their dedication to pushing boundaries and expanding what’s possible. What’s next? More frameworks, more capabilities, and plenty of Early magic to come. 💡 Check out our documentation to see how to get started with Vitest and other supported frameworks (link in the first comment 👇). #DeveloperTools #UnitTesting #GenAI #Vitest

🚨 Reminder: Node.js IL Monthly Meetup is Almost Here! 🚨 📅 When: December 22, 2024 📍 Where: Tel Aviv, Israel Join us for an evening of learning, coding, and connecting. 🗓 Agenda: 18:00 🎉 Mingling, 🎁 SWAG, 🍕 Pizza & 🍻 Beers 18:30 💻 "JavaScript is multi-threaded?!" w/ Gil Tayar 19:05 🤖 "GenAI DevTools for Unit Tests" w/ Sharon Barr (CEO, Early) 19:45 🔄 "Why Use Message Queues" w/ Dmitry Porotov 20:30 🏡 Wrap-up Looking forward to seeing you there—don’t miss out! 👉  RSVP in first comment

Building a startup is a journey of a million small steps. Today, we’re proud to share that our documentation website is live 🌐—marking another milestone toward delivering the best experience for developers. For the tech enthusiasts: Our documentation site is built using Docusaurus, hosted on GitHub Pages with a custom domain configured through our hosting provider. The project is developed in VSCode and lives on our docs GitHub repository. This setup makes it easy for us to add content, build and test locally, and publish updates seamlessly. What’s New in the Docs? We’ve completely revamped our documentation to give developers the resources they need to succeed with the EarlyAI extension. Whether you’re getting started or diving deeper into EarlyAI’s capabilities, our new site will help you onboard quickly. Here’s what you can explore: 📖 Quickstart guides 🔍 Supported languages and frameworks (TypeScript, JavaScript, and Python for now) 🧑💻 Supported test frameworks and configuration tips (Jest, Mocha, Vitest, Pytest) 💡 How to use the extension (generate and view tests) 🪄 Using Earl, your AI Test Engineer workforce 🛠️ Configuring naming conventions for your test files And this is just the beginning—more documents and features are already in the works! 👉 Explore the new docs (link in the first comment). #DeveloperTools #UnitTesting #GenAI

Fantastic feedback from an Early Adopter Python User of EarlyAI 🚀 Yesterday, a friend who tried EarlyAI for Python for the first time shared his candid experience with me. Here's what stood out: To evaluate Early, he started with code he’d already written unit tests for, and here’s what he noticed: 🔹 He typically uses the unittest framework but decided to try Early with pytest - and found it seamless. The tests were roughly similar to his, but: 🔹 Early automatically generated mocks that he had to write manually. 🔹 All tests were green (no red tests). 🔹 What took him an hour to do manually, EarlyAI did in just 8 seconds! 🔹 He’s now eager to try Early on more complex code and if we generate red tests, and what they might reveal. 🔹 He also plans to demo Early to his VPE after diving deeper into these scenarios. 💡 User Experience Feedback: 🔹 He suggested making the code lens for generating tests configurable, as it can be distracting when he’s focused on writing code and not ready to generate tests yet. It’s always inspiring to see how EarlyAI is saving developers time and helping them tackle more challenging problems. I’m looking forward to seeing how it performs on his complex cases! I welcome the developers reading this to give it a try too, I am looking for direct feedback if you are up to it. feel free to message me. #Python #UnitTesting #DeveloperExperience Here is an example of unit tests generated by Early 👇

🚀 Exciting News: Early announce Python Support for Auto Unit Test Generation! 🐍 Python is one of the top three most popular programming languages, and today, we're thrilled to share that EarlyAI extension now supports auto unit test generation for Python as well! At Early, our mission is to help developers deliver higher-quality code faster. To test and showcase Early's capabilities, we generated tests for the popular open-source project psf/requests (⭐ 52,300 GitHub stars) - a simple yet elegant HTTP library. Here's what we achieved in under an hour: ✅ 944 passing tests (green) ❌ 247 failing tests (red)—we're analyzing these to identify potential bugs. 📊 73% code coverage achieved! ⏱️ Under an hour As a reminder, EarlyAI generates validated and verified unit tests right in your IDE - in just a click. Try us out and start generating unit tests with a single click 👇 #Python #UnitTesting #OpenSource #genAI

🎙Podcast Alert - מפתחים מחוץ לקופסה! Hosted by Dotan Talitman & Shahar Polak, featuring Early’s founders Lior Froimovich & Sharon Barr, this episode was 🔥! We talked all things GenAI—from the mind-blowing to the downright hilarious: 💡 How do tools move from assist to agent? 🤔 Is OpenAI my assistant or am I theirs? ⚠️ What are the risks of letting GenAI loose on our code? 🛡️ How can we safeguard developers from AI chaos? 🚀 What’s next for developers and our entire ecosystem in 5 years? This conversation was anything but ordinary—fun, honest, and full of surprises. Don't miss it! And while you’re at it, supercharge your VSCode with EarlyAI and start generating unit tests today. Link to Podcast & Add EarlyAI to VSCode in comments below 👇

🚀 Join us for the Dec 2024 Node.js Israel Monthly Meetup! 📅 When: 22nd Dec, 2024 📍Where: Tel Aviv, Israel 18:00 - 🎉 Mingling, 🎁 SWAG, 🍕 Pizza & 🍻 Beers 18:30 - 💻 JavaScript is multi-threaded?! w/ Gil Tayar 19:05 - 🤖 GenAI DevTools for Unit Tests w/ Sharon Barr (EarlyAI CEO) 19:45 - 🔄 Why Use Message Queues w/ Dmitry Porotov 20:30- 🏡 Wrap-up Let’s code, learn, and connect with fellow devs! 📍 Details & RSVP below:

🚀 AI Assistant vs. AI Agent We've all been hearing terms like AI Assistant and AI Agent for quite some time now, and how they're transforming development workflows. But do you really know the difference? 🤔 Check out our blog to uncover what sets them apart and how EarlyAI is taking software testing to the next level! See links in comments 👇

Cursor vs. GitHub Copilot vs. EarlyAI: Which One Should You Use for Unit Test Generation? Choosing the right tool depends on the task. General-purpose tools are versatile, but when the goal is to improve the quality of your new or existing code, it’s essential to compare them for that specific purpose. Let’s take a closer look at Cursor and EarlyAI for the task of unit test generation and improving code quality - developers' ultimate goal: AI Coding Assistant (Cursor) for Unit Test Generation Cursor assists in generating test code, but developers must manually review and validate the results. The generated code often doesn’t compile or work right away, requiring significant effort to sift through suggestions one by one. This process can be high-maintenance, time-consuming, and not scalable. AI Agent (EarlyAI) for Unit Test Generation EarlyAI integrates seamlessly with tools like GitHub Copilot or Cursor, generating dozens of high-quality, functioning unit tests within minutes. The value is clear in the high test coverage and time saved. EarlyAI can generate hundreds of tests in less than an hour, making bulk test management simple. For example, EarlyAI produced 206 unit tests with 89% code coverage, for ExpressoTS. work of a month was done in a couple of hours. Moreover, EarlyAI goes a step further by generating failing tests ("red tests") that reveal bugs. 10% of the tests on the above example were failing, most of them revealed bugs in the code. This reflects EarlyAI's deep understanding of code intent - not just its implementation. Try it for yourself. Search for EarlyAI in the VSCode marketplace and generate tests in a minute.

Check out this brief summary of Early's fantastic webinar with Richard Zampieri from ExpressoTS, where he shared his experience using Early for his project. Quick summary: 206 unit tests. 89% coverage. 10% test failing, discovered issues in the code. 2 hours of work. “If I was doing this manually, I would be spending more than a month, but I was able to do this in a couple of hours" "10% of tests were failing, majority were issues through my code… I was able to spend time reviewing real bugs” Link to the full webinar is on the first comment