Claude Code Env Precheck

Audit your environment before AI writes a single line of code. Prevent dependency failures, path issues, and disk overflows before they happen.

License: MIT Claude Code

English | 简体中文

---

🌟 Why this Skill?

When using Claude Code to scaffold projects, Claude often starts creating files immediately without checking your local environment. This leads to:

• npm install failures: Debugging for hours only to find it's a WSL path issue.
• Missing Runtime: The boilerplate is done, but you don't even have Node/Python installed.
• Disk Anxiety: Not knowing that node_modules will eat up 500MB of your remaining space.

This skill forces Claude to be a responsible architect instead of just a "coder".

📸 Demo

🚀 Quick Install

The easiest way to install is to let Claude do it for you. Copy and paste this into your Claude Code terminal:

claude "Please download [https://raw.githubusercontent.com/Gray7071/claude-code-env-precheck/main/SKILL.md](https://raw.githubusercontent.com/Gray7071/claude-code-env-precheck/main/SKILL.md) and save it to my global skills as env-precheck.md"

🛠️ Key Features
Environment Auditing: Checks Node, Python, Git, and more before execution.

Disk Estimation: Provides a rough estimate of node_modules and build artifacts.

Constraint Awareness: Warns about specific environment issues like WSL UNC path limitations.

🌟 为什么需要这个 Skill？
在 Claude Code 中搭建项目时，AI 通常会直接开始创建文件，跳过环境检查。这会导致：

安装失败： 报错排查半天，结果发现是路径不兼容。

缺少运行时： 代码写完了，才发现本地没装 Python 或 Node。

存储隐患： 不知道依赖包会占用多少磁盘空间。

这个 Skill 让 Claude 在动手前先回答三个问题：需要什么？你有什么？还缺什么？

🚀 快速安装
最简单的安装方式是让 Claude 帮你自己安装。在终端输入：

Bash
claude "请下载 [https://raw.githubusercontent.com/Gray7071/claude-code-env-precheck/main/SKILL.md](https://raw.githubusercontent.com/Gray7071/claude-code-env-precheck/main/SKILL.md) 并将其另存为我的全局技能 env-precheck.md"
🛠️ 核心功能
环境审计： 在启动前自动检查运行时版本（Node, Python 等）。

空间预估： 预估依赖包大小，避免磁盘爆满。

限制提示： 自动识别 WSL 路径限制等常见坑点。

📄 License
This project is licensed under the MIT License.


---

### 💡 接下来你只需做三件事：

1.  **替换图片链接：** 把你那张照片改名为 `demo-screenshot.png`，放进 `examples` 文件夹里，然后上传。我在代码里已经帮你写好路径了。
2.  **修改链接中的用户名：** 将上面安装命令里的 `Gray7071` 确认无误。
3.  **在 GitHub 设置里添加 Tags：** 别忘了我之前提过的 `claude-code`、`ai-agents`