📄 中文摘要
在现代软件开发中，保持干净架构的完整性至关重要。通过自定义的 detekt 规则、专业化的 AI 代理和基于规范的开发方法，可以有效地防止大型语言模型（LLM）在代码生成过程中引入架构问题。作者分享了一个实例，描述了如何在 LLM 生成的代码中发现架构破坏，并通过设置构建错误来强制开发者关注架构的完整性。这种方法不仅提升了代码质量，还确保了系统的可维护性和可扩展性。
一个大型语言模型在一次生成中破坏了我的架构。我将其设为构建错误
📄 English Summary
An LLM Broke My Architecture in One Generation. I Made That a Build Error
Maintaining the integrity of Clean Architecture is crucial in modern software development. Custom detekt rules, specialized AI agents, and specification-driven development methods can effectively prevent large language models (LLMs) from introducing architectural issues during code generation. The author shares an instance where an LLM's generated code broke the architecture, and how setting a build error forced developers to focus on maintaining architectural integrity. This approach not only enhances code quality but also ensures the system's maintainability and scalability.