展示HN:我在一个周末使用AI工具构建了一个睡前故事网页应用程序
在周末,我进行了一项小实验,看看我能多快从构思到实现一个可工作的最小可行产品(MVP),使用人工智能辅助开发工具。<p>结果是一个简单的网页应用程序,它生成简短的个性化睡前故事,并用我自己的声音进行讲述。代码库在某些地方故意保持粗糙,但应用程序可以正常工作,并且在实际环境中可用。<p>让我感到最惊讶的是,能够在一个真实的、已部署的系统上轻松迭代,而不被早期的架构决策所束缚。<p>技术栈:FastAPI、Supabase、Fly.io、Vercel<br>工具:Cursor、Lovable、Claude、ChatGPT<p>我很好奇这里的其他人是如何进行人工智能辅助开发的,以及在将事物固化为生产环境之前,你们是如何划定界限的。<p>链接:<a href="https://sleepli.app" rel="nofollow">https://sleepli.app</a>
查看原文
Over the weekend I ran a small experiment to see how quickly I could go from idea → working MVP using AI-assisted development tools.<p>The result is a simple web app that generates short, personalized bedtime stories and narrates them in my own voice. The codebase is intentionally rough in places, but the app works and has been usable in a real setting.<p>What surprised me most was how easy it was to iterate on a real, deployed system without getting stuck on early architectural decisions.<p>Stack: FastAPI, Supabase, Fly.io, Vercel
Tools: Cursor, Lovable, Claude, ChatGPT<p>Curious how others here are approaching AI-assisted development, and where you draw the line before hardening things for production.<p>Link: <a href="https://sleepli.app" rel="nofollow">https://sleepli.app</a>