告别“聊天式”写代码，看Web开发者如何把AI从玩具变成生产工具
The End of 'Vibe Coding' For the past year, the industry has been enamored with 'vibe coding'—the act of throwing natural language prompts at an LLM a…
10M tokens烧出一夜全库类型注解，AI批量重构老代码的硬核实战，脚本加耐心就够。
Ten million tokens sounds infinite. It is not. I burned through a large chunk of it in one weekend. The result? A legacy Python codebase with type ann…
两千次提交背后，是反复重写和自我较劲的成长故事，生日快乐！
Today's my birthday. And today, without planning it, my SaaS WorkoutNote shipped its 18th update on the exact same day the repo crossed 2000 commits .…
吐槽Claude Code过于保守，探讨如何调教它从“避雷”型助手变成真正懂重构的工程师，HN高赞讨论。
Claude code is living in the past. Like an atomically powered steam engine. The LLM is trained on the past. The capabilities have leapt forward, but i…
拒绝空谈，五个实操技巧让AI真正融入日常开发，从任务拆解到代码重构，实用不炒作。
A Practical Starter Guide to Using AI in Your Daily Dev Workflow (Without the Hype) AI tools are everywhere, but “use AI” is vague advice. The fastest…
探讨多LLM编排框架，利用Gemini重构、GPT/Claude写代码的互补优势，解锁AI协作新思路。
Here is my general take: I feel Gemini is excellent at high-level refactoring but riddled with bugs when writing actual code. On the other hand, GPT/C…
Claude Opus因训练数据限制，无法识别新颖代码模式，陷入重构循环。
I'm having a terrible problem with claude opus constantly clobering some of my codebase ..because I'm doing some things that are novel and arent incld…
最新研究发现：LLM编码代理能否像人类一样进行代码重构？论文评估其可靠性与能力边界。
arXiv:2603.04177v2 Announce Type: replace-cross Abstract: LLM coding agents can generate working code, but their solutions often accumulate complexity…
用Copilot将老旧的Python游戏原型翻新成JavaScript版，从重构到模块化，AI辅助开发全程实战。
This is a submission for the GitHub Finish-Up-A-Thon Challenge What I Built Pixel Siege is a fast-paced 8-bit arcade shooter game where players battle…
用GitHub Copilot将组合优化器从5秒提速到1.2毫秒，开源杂货比价工具的背后故事
This is a submission for the GitHub Finish-Up-A-Thon Challenge What I Built Cijene ("Prices") is Croatia's premier open-source grocery price tracker a…
提出将LLM生成的形式化证明自动重构为可复用模块的新方法，提升证明可维护性与复用性。
arXiv:2606.03743v1 Announce Type: new Abstract: While Large Language Models (LLMs) have shown strong performance in generating formal proofs, their ou…
31万行代码AI重构实战，关键在于用Agent评测思路对齐人机共识，三个经验直击AI Coding管理痛点。
当 90% 以上代码由 AI 生成，决定系统走向的不是谁写得更快，而是约束 AI 的能力。没有统一规范，AI 只会成倍放大混乱。本文基于 31 万行代码重构实践，分享我们如何用 Agent 评测思路管理 AI Coding——通过技术债梳理、建设Rule、重构 SOP 和 Pre-PR 机制，把重构…