Token导航 LogoToken导航

Agent Skill · 开发

smart-skill-maker(智能技能制造者)

smart-skill-maker 用于处理 GitHub 仓库、Issue、Pull Request 和代码协作信息,适合在 Codex、Claude、Cursor、Gemini CLI 中需要围绕仓库状态、代码变更或协作事项进行整理时使用。可结合来源仓库、安装命令和原始 README 继续核验具体用法。安装前建议确认权限范围、维护状态,以及是否会触发联网、命令执行或文件读写。

smart-skill-maker 是开发类 Agent Skill,可用宿主为 Codex、Claude、Cursor、Gemini CLI,通过GitHub 仓库获取,权限为只读;许可证信息还不明确,第三方扫描已通过,按安装量在开发类的 21,746 个 Skill 里排第 16,191。

总安装 524周安装 21GitHub Stars 公开资料未说明下载 170核验 2026-05-01

安装说明

本站只整理中文说明,不托管安装包,也不代用户安装。
GitHub
许可证
unknown

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:smart-skill-maker(智能技能制造者)
来源仓库:https://github.com/lingkaix/smartworkers
仓库路径:skills/smart-skill-maker
安装命令:
npx skills add https://github.com/lingkaix/smartworkers --skill smart-skill-maker
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。该命令会通过 npx skills 从第三方来源获取 Skill;本站只展示命令,不托管安装包,也不自动执行。

skills.shnpx skills
npx skills add https://github.com/lingkaix/smartworkers --skill smart-skill-maker

这个 Skill 是做什么的

开发类 Skill 面向写代码、改代码和排查问题,一般会读写仓库文件,部分还会执行构建或测试命令。

smart-skill-maker 用于处理 GitHub 仓库、Issue、Pull Request 和代码协作信息,适合围绕项目变更进行整理。适用于技能创建和协作事项管理的开发场景。通过 npx skills add 命令从指定 GitHub 仓库安装并使用。

安装前需确认权限范围和维护状态,注意可能涉及文件读写操作。建议结合原始 README 了解具体支持的协作功能和操作边界。

来源仓库内路径:skills/smart-skill-maker

上手步骤

4 步
  1. 1

    确认自己的宿主工具

    公开资料标注支持 Codex、Claude、Cursor、Gemini CLI;先确认自己用的是其中之一,再按对应工具的 Skills 目录放置。

  2. 2

    拿到安装命令

    可以直接复制的命令是 npx skills add https://github.com/lingkaix/smartworkers --skill smart-skill-make…;命令里的占位符换成自己的路径或密钥再执行。

  3. 3

    安装后重启宿主

    命令执行完要完全退出宿主工具再打开,让新装的 Skill 被重新扫描到,否则列表里可能看不到。

  4. 4

    按最小权限试用

    它标注为只读,先用只读任务试一次,确认行为符合预期再扩大使用范围。

常见问题

6 条

smart-skill-maker 怎么安装?

在终端执行 npx skills add https://github.com/lingkaix/smartworkers --skill smart-skill-maker。装完完全退出宿主工具再打开,让它重新扫描本地 Skill。

smart-skill-maker 支持哪些 AI 工具?

公开资料标注支持 Codex、Claude、Cursor、Gemini CLI。其他工具只要能读取本地 Skill 目录,一般也能手动放进去,但要自己确认触发方式是否一致。

smart-skill-maker 需要什么权限?

权限级别是只读。该 Skill 主要提供规则、说明或参考内容,本身偏只读;真正读写文件、联网或执行命令仍取决于宿主 Agent 的任务。

smart-skill-maker 是免费的吗?

公开资料没有标注明确许可证,收费和商用限制都无法确认。正式使用前请到来源仓库核对授权说明。

smart-skill-maker 用的人多吗?

公开数据是累计安装 524 次,近一周安装 21 次,下载 170 次。这些数字来自第三方目录统计,只能当作热度参考,不代表质量或安全性。

smart-skill-maker 安全吗?

第三方扫描结果为通过,但这只覆盖已知的静态检查项,不能替代自己读一遍源码和安装命令。

值得选的地方和要先想清楚的地方

值得选的地方

  • 同时适配 4 种工具,换客户端不用重新找替代品。
  • 第三方扫描已通过,常见的高风险行为没有命中。
  • 标注为只读权限,试用阶段不容易误改本地文件。

先想清楚的地方

  • 许可证没有明确标注,能不能商用、能不能再分发都不确定。
  • 第三方目录整理,版本更新和行为细节以来源仓库为准。

适不适合你

更适合

  • 已经在用 Codex,想直接扩展它能力的人。
  • 需要处理开发类任务、又不想每次都手写提示词的人。
  • 只想先试试效果、不希望它改动本地文件的人。

不太适合

  • 需要明确授权才能用到公司项目里的场景。
  • 需要厂商级技术支持和 SLA 的生产系统。

开发类 Skill 对比

6 个同类
Skill总安装GitHub 星权限许可证扫描
skill-vetter

技能审查员 · OpenClaw

8,779,018986操作浏览器MIT-0通过
ontology

本体论 · OpenClaw

6,636,960576只读MIT-0通过
entra-app-registration

内部应用程序注册 · Codex、Claude

6,631,872764接触敏感数据未明确通过
entra-app-registration

内部应用程序注册 · Codex、Claude

3,757,680201接触敏感数据未明确通过
nano-banana-pro

纳米香蕉专业版 · OpenClaw

3,526,898364接触敏感数据MIT-0有提醒
smart-skill-maker

智能技能制造者 · 本页

524—只读未明确通过

按安装量取同分类前几名;数字来自第三方目录统计,只作热度参考,不代表质量结论。

SKILL.md 原文

Smart Skill Maker

This is the single SmartWorkers skill-authoring entry point.

It uses $skill-creator as the core drafting and improvement engine, then applies the repo's shared conventions from:

  • skills/general/smart-skill-maker/assets/templates/SKILL.md
  • skills/general/smart-skill-maker/references/AUTHORING.md
  • skills/general/smart-skill-maker/scripts/new_skill.sh

Required inputs

  • Target role and skill name
  • Whether the task is creating a new skill or improving an existing one
  • The workspace root path
  • Whether the source copy already exists under skills/<role>/<skill-name>/
  • Whether the skill should also include agents/openai.yaml
  • The worker role the skill belongs to

Workflow

  1. Confirm the workspace authoring prerequisites.

- skill-creator should already be installed for Codex. - npx skills should already be available. - The workspace should have a local skills/ folder with skills/AGENTS.md.

  1. Load the canonical SmartWorkers authoring materials.

- Use skills/general/smart-skill-maker/assets/templates/SKILL.md as the default scaffold. - Use skills/general/smart-skill-maker/references/AUTHORING.md as the house-style reference. - Keep skills/AGENTS.md short and operational: it should define local apply behavior, not duplicate the full style guide.

  1. Make skills/ the source of truth.

- For a new skill, create skills/<role>/<skill-name>/. - Treat <role> as the worker identity and usage scope. A role is not just a tool category; it represents a real worker with related skills and one or more workflows. - Similar capabilities may appear in different roles when the real job, workflow, or acceptance criteria differ. For example, a copy writer and an ads designer may both have image-generation skills, but those skills should still stay role-specific if the worker expectations are different. - For an existing installed skill with no source copy yet, copy .agents/skills/<skill-name>/ into skills/<role>/<skill-name>/ first, using the role that worker should own. - After that point, edit only the source copy in skills/.

  1. Use $skill-creator for the draft or improvement loop.

- Let $skill-creator handle generic skill drafting, refinement, and iteration. - Keep SmartWorkers-specific structure here: - routing detail in frontmatter description - safe placeholders in workers.example.jsonc - working files in logs/ - final deliverables in artifacts/ - reusable helpers in scripts/, assets/, and references/

  1. Apply SmartWorkers conventions before finishing.

- Keep name aligned with the skill folder name, and keep that name stable across routine revisions. - Add and maintain a frontmatter metadata.skill_version value so the skill revision is easy to identify without renaming the skill. - Increment metadata.skill_version every time the skill is updated. - Keep the skill body lean and operational. - Keep agents/openai.yaml in sync when present. - Keep documented file references repo-relative; do not hardcode machine-specific absolute filesystem paths. - Avoid secrets and real credentials in skill files. - When the authoring workflow changes, update this skill's assets/, references/, and scripts/ so the installed workflow and the written conventions do not drift apart.

  1. Scaffold safely when a new source folder is needed.

- From the repo root: - bash skills/general/smart-skill-maker/scripts/new_skill.sh --role <role> --name <skill-name> - Preview output goes to logs/smart-skill-maker/<timestamp>/<role>/<skill-name>/. - Use --apply only after reviewing the scaffold and making any SmartWorkers-specific adjustments.

  1. Apply the source skill to Codex with npx skills.

- From the workspace root, run: - npx skills add -a codex -y./skills --skill <skill-name> - Use npx skills as the only install/update path into .agents/skills/. - Do not manually patch .agents/skills/<skill-name>/.

  1. Report the result.

- Name the source folder under skills/. - Confirm the apply command used. - Confirm the installed Codex copy is now managed by npx skills.

Outputs

  • Source skill under skills/<role>/<skill-name>/
  • Installed Codex skill managed under .agents/skills/<skill-name>/
  • Shared templates and authoring guide under skills/general/smart-skill-maker/
  • Any working notes or review material under logs/

Definition of done

  • The source copy exists under skills/<role>/<skill-name>/
  • The skill follows SmartWorkers conventions
  • The skill keeps a stable name and an explicit frontmatter metadata.skill_version
  • metadata.skill_version was incremented for this update
  • The updated skill has been applied to Codex with npx skills add -a codex -y./skills --skill <skill-name>
  • No manual source-of-truth edits were left only inside .agents/skills/
  • If the authoring workflow changed, the supporting assets/, references/, and scripts/ under skills/general/smart-skill-maker/ were updated too

Safety / quality checklist

  • Do not treat .agents/skills/ as the editable source of truth.
  • Do not mix manual copying into .agents/skills/ with npx skills updates.
  • Do not commit secrets or paste credential values into skill files.
  • Do not let the installed smart-skill-maker contract drift away from the shared template and authoring guide.

平台分布

Codex

33.42%
按下载量换算57

Claude

32.15%
按下载量换算55

Cursor

18.93%
按下载量换算32

Gemini CLI

10.34%
按下载量换算18

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

通过

权限和风险

只读

该 Skill 主要提供规则、说明或参考内容,本身偏只读;真正读写文件、联网或执行命令仍取决于宿主 Agent 的任务。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。目前只有一个公开来源,建议同时核对源仓库。

继续浏览同类 Skills