Token导航 LogoToken导航

Agent Skill · 开发

align-repo(对齐回购协议)

align-repo 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。

align-repo 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,第三方扫描有提醒项,按安装量在开发类的 21,746 个 Skill 里排第 17,643。

总安装 392周安装 16GitHub Stars 4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:align-repo(对齐回购协议)
来源仓库:https://github.com/apocohq/skills
仓库路径:skills/align-repo
安装命令:
npx skills add https://github.com/apocohq/skills --skill align-repo
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/apocohq/skills --skill align-repo

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。

对齐回购协议用于审计仓库配置是否符合团队标准,检查 .claude/settings.json 和 CLAUDE.md 设置。适合团队协作环境中确保开发规范一致性的场景,支持 Codex、Claude 等主流宿主。按清单逐项检查配置对齐状态,对不合规项提出修改建议但不会自动应用变更。

安装命令:npx skills add https://github.com/apocohq/skills --skill align-repo 执行时会严格逐步提问,每次只处理一项,需用户明确授权后才能进行修改操作。

来源仓库内路径:skills/align-repo

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/apocohq/skills --skill align-repo;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

align-repo 怎么安装?

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

align-repo 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。

align-repo 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。

align-repo 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。

align-repo 用的人多吗?

公开数据是累计安装 392 次,近一周安装 16 次,下载 125 次,GitHub 4 星。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。

align-repo 安全吗?

第三方扫描有提醒项,说明存在需要人工确认的行为。安装前请打开源码看清它读了什么、联网到哪里。

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。

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。
  • 第三方扫描存在提醒项或未通过项,需要人工逐条确认。
  • 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
align-repo

对齐回购协议 · 本页

3924需要联网未明确有提醒

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

Align Repo Conventions

Audit the current repo's .claude/settings.json and CLAUDE.md against team standards.

Execution Rules

  • Execute the checklist strictly step by step.
  • Present ONE item per message. Show its status, and if action is needed, ask and STOP.
  • Do not present the next item until the user has responded.
  • If an item is already aligned, say so and immediately proceed to the next item in the same message.
  • If an item needs action, show current vs expected and ask: *"Want me to align this?"*
  • Never batch multiple action items into one message.
  • Never auto-apply changes — every modification requires explicit user confirmation.
  • Merge, don't overwrite — add missing keys without clobbering existing ones.
  • Preserve existing content in CLAUDE.md — don't remove or reorder existing sections.

Checklist

  • 0. Self-update — Run: npx skills add https://github.com/apocohq/skills --skill align-repo -a claude-code -y. If updated, tell the user: *"align-repo was updated. Please start a new session and re-run /align-repo to use the latest version."* Stop — do not continue with outdated instructions.
  • 1. Gather state — Read .claude/settings.json, CLAUDE.md, and check for .claude/settings.local.json (note existence, don't modify). If .claude/settings.json doesn't exist, ask if you should create it. If CLAUDE.md doesn't exist, ask if you should create it.
  • 2. Attribution config — Check .claude/settings.json for the expected attribution block (see Reference A).
  • 3. Allowed tools — Check allowedTools as a set (see Reference B). Report missing entries and flag extra entries (don't remove extras — they may be project-specific).
  • 4. Worktrees convention — Check .gitignore includes .worktrees/. Check CLAUDE.md for a Worktrees section (see Reference C).
  • 5. Architecture principles — Check CLAUDE.md for a Separation of Concerns & DRY section (see Reference D).
  • 6. Commit conventions — Check CLAUDE.md for a Commit Conventions section covering all 5 items (see Reference E).
  • 7. Meeting format — Skip if no meetings/ directory. Check CLAUDE.md for a Meeting Format section (see Reference F).
  • 8. Recommend skills — Review repo purpose/tech stack. For each relevant skill from Reference G, explain why it fits and ask to install. Skip skills that don't match.
  • 9. Check skill updates — Run npx skills check. If updates available, ask to run npx skills update. Skip if no skills-lock.json.
  • 10. Summary — Print the alignment summary (see Reference H).

References

A. Expected attribution

{
  "attribution": {
    "commit": "Assisted-By: Claude (Anthropic AI) <noreply@anthropic.com>",
    "pr": ""
  }
}

B. Expected allowed tools

{
  "allowedTools": [
    "Bash(git status*)",
    "Bash(git diff*)",
    "Bash(git log*)",
    "Bash(git fetch*)",
    "Bash(git branch*)",
    "Bash(git checkout -b *)",
    "Bash(git stash*)",
    "Bash(git add *)",
    "Bash(git commit *)",
    "Bash(gh issue view*)",
    "Bash(gh issue list*)",
    "Bash(gh pr view*)",
    "Bash(gh pr list*)",
    "Bash(ls:*)"
  ]
}

C. Expected worktrees convention

## Worktrees

Use `.worktrees/` for git worktrees. Branch naming follows commit conventions (e.g., `feat/session-history`).

### Setup

After creating a worktree, run project setup:

- **Node.js**: `pnpm install`
- **Python**: `uv sync`

### Verification

Run tests to confirm a clean baseline before starting work. If tests fail, report failures and ask before proceeding.

### Report

After setup, report: worktree path, test results, and readiness.

D. Expected architecture principles

## Separation of Concerns & DRY Principle

This system is a modular component system following the DRY (Don't Repeat Yourself) principle. Each piece has a single responsibility. You should be able to swap out any component without rewriting others.

E. Expected commit conventions

## Commit Conventions

- **Conventional Commits**: `type(scope): short summary` — types: `feat`, `fix`, `docs`, `refactor`, `test`, `chore`, `revert`, `style`, `perf`, `ci`, `build`.
- **Scope**: Optional but encouraged (e.g., `feat(ui):`, `fix(hook):`, `docs(design):`).
- **Body**: Optional concise bullet points for non-trivial changes.
- **Trailer**: Configured via `.claude/settings.json` `attribution` — do not add manually.
- **Branch naming**: `type/short-description` (e.g., `feat/session-history`, `fix/stale-timer`). Same type prefixes as commits.

F. Expected meeting format

## Meeting Format

---
date: "YYYY-MM-DD"
attendees:
  - First Last
  - First Last
---

# Meeting Title

One concise paragraph summarizing the meeting.

## Transcript / Notes / Meeting Minutes

G. Available skills

SkillWhat it doesWhen to suggest
process-transcriptConverts VTT meeting transcripts into structured markdown notesRepos that track meetings or have a meetings/ directory
ralph-itPicks the next user story from a PRD GitHub issue, implements it, and opens a PRRepos that use GitHub issues for PRDs or have structured user stories
write-a-prdInterviews the user and writes a PRD, then submits it as a GitHub issueAny repo that plans features via PRDs or GitHub issues
review-workDispatches a code-reviewer subagent to review changes before pushingAny repo where code quality reviews before push are desired
adrCreates, lists, and updates Architecture Decision Records in docs/adrs/Repos that track architectural decisions or have a docs/adrs/ directory

Install command: npx skills add https://github.com/apocohq/skills --skill <skill-name> -a claude-code -y

H. Summary template

## Alignment Summary
- settings.json attribution: [aligned / changed / skipped]
- settings.json allowed tools: [aligned / changed / skipped]
- .gitignore worktrees: [aligned / added / skipped]
- CLAUDE.md worktrees: [present / added / skipped]
- CLAUDE.md architecture principles: [present / added / skipped]
- CLAUDE.md commit conventions: [present / added / skipped]
- CLAUDE.md meeting format: [present / added / skipped / n/a]
- Skills installed: [list or "none"]
- Skills updated: [list or "all up to date"]
- Items skipped by user: [list if any]

平台分布

Codex

37.83%
按下载量换算47

Claude

30.23%
按下载量换算38

Cursor

19.81%
按下载量换算25

Gemini CLI

9.55%
按下载量换算12

安全审计

Gen Agent Trust Hub

通过

Socket

可疑

Snyk

可疑

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。来源安全扫描存在提醒或未通过项,安装前请自行复核。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