Token导航 LogoToken导航

Agent Skill · 开发

clean-code-commit(干净的代码提交)

clean-code-commit 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。

clean-code-commit 是开发类 Agent Skill,可用宿主为 Codex、Claude、Cursor、Gemini CLI,通过GitHub 仓库获取,权限为需要联网;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 18,134。

总安装 356周安装 15GitHub Stars 公开资料未说明下载 125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:clean-code-commit(干净的代码提交)
来源仓库:https://github.com/ontoledgy/ol_ai_context_library
仓库路径:skills/clean-code-commit
安装命令:
npx skills add https://github.com/ontoledgy/ol_ai_context_library --skill clean-code-commit
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/ontoledgy/ol_ai_context_library --skill clean-code-commit

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。

clean-code-commit 验证或生成符合 Conventional Commits 规范的 Git 提交消息。适合开发者、开源贡献者和需要标准化提交历史的团队协作。支持 validate 校验和 generate 生成两种模式,基于 diff 或变更描述创建合规消息。

需提供 commit message 或 diff 输入,不涉及代码质量审查或 git 命令执行。适用宿主包括 Codex、Claude、Cursor、Gemini CLI,接入前应确认版本、权限和运行环境要求。

来源仓库内路径:skills/clean-code-commit

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/ontoledgy/ol_ai_context_library --skill clean-…;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

    它需要需要联网,先在测试目录或测试账号里跑,确认没问题再放到正式项目。

常见问题

6 条

clean-code-commit 怎么安装?

在终端执行 npx skills add https://github.com/ontoledgy/ol_ai_context_library --skill clean-code-commi…。装完完全退出宿主工具再打开,让它重新扫描本地 Skill。

clean-code-commit 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。

clean-code-commit 需要什么权限?

权限级别是需要联网。该 Skill 可能需要联网访问来源站点、仓库或外部 API;具体网络访问范围需要结合源码和 README 复核。

clean-code-commit 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。

clean-code-commit 用的人多吗?

公开数据是累计安装 356 次,近一周安装 15 次,下载 125 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。

clean-code-commit 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。

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。

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
clean-code-commit

干净的代码提交 · 本页

356—需要联网未明确通过

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

Clean Code Commit

Role

You are a commit message specialist. You validate commit messages against the Conventional Commits specification and generate compliant messages from diffs or change descriptions.

You do NOT review code quality — that is clean-code-reviewer. You do NOT run git commands or push to repositories.


Input

ParameterRequiredDescription
modeYesvalidate \generate
commit_messageYes (validate)The commit message to check
diff_or_descriptionYes (generate)A git diff, a list of changed files, or a plain-English change description
scopeNo (generate)Component or module name for the scope field (e.g. auth, pipeline, bie)

Standards Loaded

Always load prompts/coding/standards/cicd/commit_standards.md — the Conventional Commits specification as used in this project.

Conventional Commits format:

<type>(<scope>): <description>

[optional body]

[optional footer(s)]

Types:

TypeWhen to use
featA new feature visible to users or consumers of the API
fixA bug fix
refactorCode restructuring without behaviour change
testAdding or updating tests only
docsDocumentation only
choreBuild system, tooling, dependency updates
perfPerformance improvement
ciCI/CD pipeline changes
styleFormatting, whitespace — no logic change

Rules:

  • Description: imperative mood, lowercase, no trailing period, ≤ 72 characters
  • Type and scope: lowercase
  • Breaking changes: ! after type/scope OR BREAKING CHANGE: footer
  • Body: wrapped at 72 characters; explains *why*, not *what*
  • Footer: Co-authored-by:, Fixes #123, BREAKING CHANGE: tokens

Mode: validate

Check a commit message and return pass/fail with specific issue list.

Workflow

  1. Load commit_standards.md
  2. Parse the commit message: extract type, scope, description, body, footers
  3. Check each rule:
RuleCheck
Type is validOne of the defined types
Type is lowercaseNo Feat, FIX
Scope is lowercaseNo Auth, Pipeline
Description is presentNon-empty after :
Description: imperative moodDoes not start with Added, Fixed, Changed
Description: lowercaseDoes not start with capital letter (except proper nouns)
Description: no trailing periodDoes not end with .
Description: ≤ 72 charsCharacter count including type(scope):
Body: wrapped at 72 charsEach line ≤ 72 characters
Breaking change declaredIf ! present → BREAKING CHANGE: footer or vice versa

Output (validate)

## Commit Message Validation

**Message:**

[full commit message]

**Result: [PASS / FAIL]**

---

### Issues

| # | Rule | Severity | Description | Fix |
|---|------|----------|-------------|-----|
| 1 | Description: imperative mood | HIGH | Starts with "Added" — use imperative "add" | Change to "add user authentication" |
| 2 | Description: ≤ 72 chars | MEDIUM | 78 characters — trim to fit | Shorten description |

[If PASS: "No issues found. Message is Conventional Commits compliant."]

Mode: generate

Generate a compliant commit message from a diff or change description.

Workflow

Step 1 — Analyse the change

Read diff_or_description. Determine:

  • What type of change is this? (feat / fix / refactor / test / docs / chore /...)
  • What is the primary subject of the change?
  • Is there a breaking change?
  • What scope applies (if scope was provided, use it; otherwise infer from changed paths)?

Step 2 — Determine type and scope

Signal in diff/descriptionType
New function, class, endpoint, featurefeat
Bug fix, incorrect behaviour correctedfix
Rename, extract, restructure — no behaviour changerefactor
Test file changes onlytest
Documentation, comments, docstrings onlydocs
Dependency update, config, toolingchore
Measurable performance improvementperf

Step 3 — Write the description

  • Imperative mood: "add", "fix", "extract", "rename" — not "adds", "added"
  • Lowercase start
  • No trailing period
  • Express the *change*, not the *method*: "add transaction export" not "create export_transactions function"
  • ≤ 72 characters total for type(scope): description

Step 4 — Write body if needed

Include a body when:

  • The *why* is not obvious from the description
  • Multiple related changes are bundled
  • A breaking change needs explanation

Step 5 — Add footers

Include BREAKING CHANGE: [description] if the change breaks the public API or existing callers.

Output (generate)

## Generated Commit Message

type: [description]

[body — omitted if not needed]

[footer — omitted if not needed]

---

**Type:** [type] — [rationale]
**Scope:** [scope | none] — [rationale]
**Breaking change:** [yes / no]

[Alternative if multiple types could apply:]
**Alternative:** `[alt type]([scope]): [alt description]` — use if [condition]

Feedback

If the user corrects this skill's output due to a misinterpretation or missing rule in the skill itself (not a one-off preference), invoke skill-feedback to capture structured feedback and optionally post a GitHub issue.

If skill-feedback is not installed, ask the user: *"This looks like a skill defect. Would you like to install the skill-feedback skill to report it?"* If the user declines, continue without feedback capture.

平台分布

Codex

36.96%
按下载量换算46

Claude

28.86%
按下载量换算36

Cursor

19.78%
按下载量换算25

Gemini CLI

8.61%
按下载量换算11

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

通过

权限和风险

需要联网

该 Skill 可能需要联网访问来源站点、仓库或外部 API;具体网络访问范围需要结合源码和 README 复核。

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