by yibie
Claude 官方最佳实践全集:173 篇中文译文 + 100 个概念的 LLM Wiki 知识网 + 交互式知识图谱/思维导图
# Add to your Claude Code skills
git clone https://github.com/yibie/claude-best-practices-zhGuides for using ai agents skills like claude-best-practices-zh.
See how claude-best-practices-zh compares with popular alternatives.
claude-best-practices-zh is an open-source ai agents skill for AI coding assistants such as Claude Code, Codex CLI, and ChatGPT, built by yibie. Claude 官方最佳实践全集:173 篇中文译文 + 100 个概念的 LLM Wiki 知识网 + 交互式知识图谱/思维导图. It has 63 GitHub stars.
claude-best-practices-zh's catalog security scan is still queued. You can run an instant dependency and prompt-injection check now with the "Scan for vulnerabilities" button above.
Clone the repository with "git clone https://github.com/yibie/claude-best-practices-zh" and add it to your Claude Code skills directory (see the Installation section above).
claude-best-practices-zh is primarily written in HTML. It is open-source under yibie on GitHub, so you can review or fork the full source.
Yes. SkillsLLM lists many other AI Agents skills you can browse and compare side by side. Open the AI Agents category from the badge at the top of this page, or use the Related Skills and comparison links further down to weigh claude-best-practices-zh against similar tools.
No comments yet. Be the first to share your thoughts!
Available after the catalog security scan passes.
⚠️ Third-Party Software Notice
This skill is third-party open-source software developed and hosted independently on GitHub. SkillsLLM lists this project for discovery and does not control or maintain the underlying repository.
Any security checks, ratings, or warnings displayed by SkillsLLM are automated and limited in scope. They do not constitute a security certification or guarantee that the software is safe, error-free, or free from malicious code, vulnerabilities, compromised dependencies, or prompt-injection risks.
Review the source code, permissions, dependencies, and configuration before installing or running any third-party skill. Use is at your own risk. To the maximum extent permitted by applicable law, SkillsLLM is not liable for losses arising from third-party software.
The deep catalog scan for this skill is still queued. Run an instant dependency check now instead.
See comparison
把 claude.com/resources/best-practices 上的全部资源抓取下来,全文译成中文,再按 Karpathy LLM Wiki 的思路编译成一套带双向链接的概念维基,并生成可交互的思维导图与知识图谱。
🌐 站点:https://claude-best-practices-zh.vercel.app
站点提供首页、概念索引、译文索引、知识图谱与思维导图,支持 自动/浅色/深色 三种主题;条目正文一律跳回本仓库阅读(RENDER_DOCS=1 node scripts/build-site.js 可切换成站内全文渲染)。
内容经 DeepSeek 三遍审校:① 逐篇对照原文修订译文(526 处)② 重新判定概念关系方向、删除噪声、补充关键缺失(141 条关系)③ 重做知识图谱可视化(分类聚类 + 力导向)。
| 层 | 目录 | 内容 |
|---|---|---|
| 原文快照 | raw/ |
抓取当天的英文 Markdown 原文,不可变,可对照校验 |
| 中文译文 | docs/zh/ |
173 篇完整译文,约 95 万字,保留全部代码块、表格、图片与链接 |
| 概念维基 | wiki/ |
100 个原子概念条目,双向链接 + Mermaid 知识图谱 + Sources 溯源账本 |
| 思维导图 | mindmap/ |
交互图谱、思维导图、Mermaid 源码 |
入口
wiki/INDEX.md 进入概念条目(GitHub 阅读)docs/README.md 的目录里按日期/产品找raw/ 里看抓取当天的原文快照每个概念条目都遵循同一套结构,便于增量维护:
wiki/子代理.md
├── front matter # 分类、英文名、别名、被发现次数
├── 摘要 # 这个概念是什么、为什么重要
├── 关键做法 # 可执行要点
├── 在 Claude 里的用法 # 落到具体产品与功能
├── 常见误区
├── 相关概念 # 双向链接(自动回填「反向链接」)
├── 知识图谱 # Mermaid 局部关系图
└── 来源(Sources) # 该概念出自哪些文章 + 原文链接
raw/ 中的原文[ADD],修正记 [FIX],见 CHANGELOG.mdmindmap/graph.json.
├── raw/ # 英文原文快照(articles / guides / webinars)
├── docs/
│ ├── README.md # 译文总目录
│ └── zh/ # 中文译文,与 raw/ 一一对应
├── wiki/ # 概念维基(INDEX.md + 100 个条目)
├── mindmap/
│ ├── graph.html # 交互式知识图谱(vis-network)
│ ├── mindmap.html # 思维导图(markmap)
│ ├── mindmap.md # 思维导图源文件
│ ├── knowledge-graph.mmd # Mermaid 全图
│ └── graph.json # 概念与关系数据
├── data/ # 中间数据:manifest、概念卡、注册表、图谱
├── scripts/ # 抓取 / 翻译 / 概念抽取 / 建图 / 出站脚本
└── index.html # GitHub Pages 入口
npm install # playwright-core / turndown / jsdom
node scripts/manifest.js # 1. 抓资源清单(claude.com/api/resources/category)
node scripts/fetch.js # 2. 抓正文 → raw/*.md(幂等,已存在则跳过)
node scripts/translate.js # 3. 翻译 → docs/zh/*.md(断点续跑)
node scripts/concepts.js # 4. 每篇文章抽概念卡 → data/cards/
node scripts/taxonomy.js # 5. 合并同义概念、归类 → data/registry.json
node scripts/graph.js # 6. 汇总节点与关系 → data/graph.json
node scripts/wiki.js # 7. 写概念条目 → wiki/*.md
node scripts/fixlinks.js # 8. 双向链接 + 反向链接回填
node scripts/site.js # 9. 索引、思维导图、图谱 HTML
# 审校(deepseek-v4-pro)
node scripts/qa-translate.js # 10. 逐篇对照原文找翻译错误
node scripts/qa-apply.js # 11. 应用修订(带守卫:唯一匹配才改)
node scripts/qa-associations.js # 12. 审查概念关系(删除/反转/补充)
node scripts/apply-associations.js
node scripts/relink.js # 13. 用新关系重建「相关概念」与 Mermaid
node scripts/graph-view.js # 14. 生成聚类力导向图谱
# 站点
node scripts/site.js # 15. 索引 + 思维导图源文件
node scripts/build-site.js # 16. 站点产物 → site/(首页取 site-src/landing.html)
vercel --prod # 17. 发布到 Vercel
翻译与概念抽取走 DeepSeek API,读取 ~/.pi/agent/auth.json 中的密钥;所有脚本都是幂等 + 断点续跑,新增文章时重跑即可。
说明:claude.com 页面标注「232 条资源」,分页 API 实际返回 231 条(有 1 条未出现在任何分页中),本仓库以 API 返回结果为准。
NOTICE.md。最后构建:见 CHANGELOG.md | 数据源:claude.com/resources/best-practices