zerostep

d flakiness, and avoid constant rewrites when implementation changes. This is the first AI-driven application I've built. I was inspired by a couple of things: * SWE-agent's [1] focus on agentic performance * backend-GPT's README rant [2]: "The proper format for business logic is human intelligence." * zerostep [3], autogpt [4], and other browser-controlled AI projects Potential Next Steps: * Save passing specs as playwright code and only fallback on spec failure to the AI agent, to see if the test can be self-healed according to the original spec. * Create a curated benchmark of both common open source web apps that should pass, and versions with introduced bugs * NPM package to run like `npx autospec --url https://example.com` * Github action and Vercel checks API integration to run on preview deployments * Handling app auth securely+easily * Continue exploring the balance between vision UI interpretation and DOM analysis [1]: https://github.com/princeton-nlp/SWE-agent [2]: https://github.com/RootbeerComputer/backend-GPT [3]: https://github.com/zerostep-ai/zerostep [4]: https://github.com/Significant-Gravitas/AutoGPT Thanks! Zach

y the time I got to respond everyone already had their pitchforks out. Bursting out a few responses in panic probably didn't help either. Anyway, that's the context. The projects in question are https://github.com/zerostep-ai/zerostep and https://github.com/lucgagan/auto-playwright. I added a section to the documentation using every concrete piece of evidence I could think of to prove that these projects are as distinct as they possibly can be given the narrow scope of the problem and possible ways to implement it (https://github.com/lucgagan/auto-playwright#addressing-controversy). I also think that, in retrospect, it would have been nice of me to acknowledge that their project was a source of inspiration for Auto Playwright. However, at no point did I copy their project, so the thought of LICENSE attribution did not even cross my mind. I am trying to be cool head about this and think of what is the proper response. More importantly, I am trying to get a perspective from folks outside of the original debate to see if I am wrong with my actions. Is copying parts of README enough to call for attribution to the original project? Is copying error message enough to call for attribution of the original project? Where is the line? I still think this whole think exploded because I effectively released something that provides comparable functionality to a commercial solution that they are working and thats what spurred this debacle. However, I also acknowledge that I am struggling to reconcile dozens of people calling me "** code thief" in private messages and my sense of righteousness, which is why I would appreciate the insights.