Claude Skill

LearnPrompt/luban-skill

Luban is an open-source skill-polishing workshop that refines usable AI skills into loadable, transmittable, verifiable, and evolvable public assets through a five-step process.

Overview

Stars378
Forks64
LanguageShell
Last pushed2026-06-12
Last synced2026-06-17
View on GitHub

Repository

OwnerLearnPrompt
Repositoryluban-skill
Full nameLearnPrompt/luban-skill
Repo ID1,266,009,367

Install this Skill

npx skills add LearnPrompt/luban-skill -g

Registry

Typeclaude_skill
Quality score70/100
Verificationreadme_parsed
Last verified2026-06-15
Platforms
Claude
Detected files
README.md

Summary

Luban is a community-driven skill-polishing workshop that transforms usable AI skills into loadable, transmittable, verifiable, and evolvable public assets through a rigorous five-step refinement process.

Chinese description

鲁班 | Luban — 将“可用的Skill”打磨成“可装载、可传播、可验证、可进化”的公共资产。Agent技能精炼工坊:验料·访行·过尺·慢刨·回炉

Key features

  • Five-step polishing workshop: material verification, field visit, sizing, slow carving, and re-smelting
  • Transforms usable skills into loadable, transmittable, verifiable, and evolvable public assets
  • Community-driven open-source collaboration for skill refinement
  • Focused on Agent skill quality and reusability

Use cases

  • Polishing raw AI skills into production-ready public assets
  • Collaborative skill verification and quality assurance
  • Building a shared library of high-quality Agent skills
  • Standardizing skill development workflows

README excerpt

<div align="center"> # 鲁班 | Luban > 把你的Skill拿到班门前,让祖师爷重新打磨一遍。 [![Agent Skills](https://img.shields.io/badge/Agent%20Skills-luban-blueviolet)](skills/luban/SKILL.md) [![skills.sh](https://skills.sh/b/LearnPrompt/luban-skill)](https://skills.sh/LearnPrompt/luban-skill) [![实战案例](https://img.shields.io/badge/%E5%AE%9E%E6%88%98%E6%A1%88%E4%BE%8B-ai--news--radar%20v0.7.0-green)](skills/luban/examples/ai-news-radar-case.md) [![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](LICENSE) **把一个"能用的Skill",打磨成"能被理解、能被安装、能被传播、能被验证、能持续进化"的公共资产。** [安装](#快速开始) · [五个动作](#五个动作) · [实战战绩](#实战战绩) · [它和同类有什么不同](#它和同类有什么不同) · [安全边界](#安全边界) · [致谢](#方法论致谢) </div> --- ## 它解决什么问题 你写了一个Skill,自己用着挺好。然后呢? - 发到GitHub,没人装——别人看不懂它是干嘛的; - README像工程说明书,没有第一屏钩子,没有能截图的产物; - 你说它"效果不错",但拿不出一个能复现的证据; - 想优化,又不知道该先动哪里——改触发词?重写工作流?补showcase? 普通的做法是"帮我润色一下"。鲁班的做法是把它当**作品**收进工坊:先挑战它值不值得雕,再看同行凭什么立足,三把尺量出短板,一刀一刀刨,每刀都要过验证门,最后发版立规矩。 ## 它会交付什么 - **一份13节的《打磨报告》**:验料结论、同行对标表(全部带URL)、生态位判断、评分表、三个打磨方向、可直接替换的改写片段 - **一张可截图的"出师证书"**:打磨前后分数、一句话新定位、绝活、下一步 - **沉淀进你仓库的验证资产**:一次性的对比脚本固化成工具,判断标准立成项目明文规矩 ## 五个动作 | 动作 | 干什么 | 一句狠话 | |---|---|---| | **验料** | 先挑战这个Skill的前提是否成立 | 朽木不可雕也,不值得就直说 | | **访行** | 联网找同行,看清生态位 | 闭门造车出不了好工具 | | **过尺** | 结构、实测、活体三把尺量分 | 绿色的CI会撒谎,要拉真实产物对账 | | **慢刨** | 冻结基线,改动过验证门才保留 | 量不过就回刀,绝不为显得干了活而多刨 | | **回炉** | 发布后留对标观察清单,下轮从反馈进 | 交活不是终点 | ## 快速开始 ![luban demo](assets/demo.gif) ```bash npx skills add LearnPrompt/luban-skill -g ``` Claude Code 用户也可以走 plugin marketplace(自动更新): ```text /plugin marketplace add LearnPrompt/luban-skill /plugin install luban ``` 装完对Agent说: ```text 让鲁班看看我这个skill:[你的Skill目录 / GitHub仓库链接 / SKILL.md内容] ``` 鲁班会先完成验料、访行、定位、过尺,给你三个打磨方向并推荐一个——**在你选方向之前,它不会动你一行字**。 ## 触发方式 - "让鲁班看看这个skill" / "班门打磨一下" - "升级我的skill" / "打磨我的skill" - "skill体检" / "skill审计

Topics

No topics yet.

Explore more

Data from GitHub. Synced on 2026-06-17