📄 中文摘要
随着 AI 编程工具的普及，开发者们越来越依赖于 GitHub Copilot、ChatGPT 和 Claude 等工具来加速代码编写。然而，AI 生成的代码存在安全隐患，研究显示其安全漏洞数量是人类编写代码的 2.74 倍。此外，经验丰富的开发者在使用 AI 工具处理复杂任务时速度降低了 19%。现有的代码审查也无法有效识别 AI 特有的模式，如虚假导入和过度注释的模板代码。为了解决这一问题，作者开发了一个工具，旨在回答“这个代码库中有多少是 AI 编写的”这一关键问题。
我构建了一个命令行工具来评估你的代码中有多少是由 AI 编写的
出处: I Built a CLI That Scores How Much of Your Code Was Written by AI
发布: 2026年2月23日
📄 English Summary
I Built a CLI That Scores How Much of Your Code Was Written by AI
As AI coding tools become ubiquitous, developers increasingly rely on tools like GitHub Copilot, ChatGPT, and Claude to accelerate code writing. However, AI-generated code poses security risks, with studies indicating that it has 2.74 times more security vulnerabilities than human-written code. Additionally, experienced developers are 19% slower when using AI tools for complex tasks. Existing code reviews also fail to effectively identify AI-specific patterns, such as hallucinated imports and over-commented boilerplate code. To address this issue, the author developed a tool aimed at answering the crucial question, 'How much of this codebase was AI-generated?'