Token导航 LogoToken导航

Agent Skill · 开发

pavlo-agent-skill-creator(巴甫洛特工技能创建器)

pavlo-agent-skill-creator 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。

pavlo-agent-skill-creator 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 19,580。

总安装 267周安装 11GitHub Stars 公开资料未说明下载 87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:pavlo-agent-skill-creator(巴甫洛特工技能创建器)
来源仓库:https://github.com/pavloglushko/ai-skills
仓库路径:skills/pavlo-agent-skill-creator
安装命令:
npx skills add https://github.com/pavloglushko/ai-skills --skill pavlo-agent-skill-creator
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/pavloglushko/ai-skills --skill pavlo-agent-skill-creator

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。

pavlo-agent-skill-creator 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 中围绕仓库状态或协作事项进行整理。通过 npx 命令从指定仓库安装,需结合原始 README 确认具体用法。

使用前应核实权限范围、维护状态及是否涉及联网或命令执行。建议优先参考来源仓库中的 SKILL.md 获取准确功能说明。

来源仓库内路径:skills/pavlo-agent-skill-creator

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/pavloglushko/ai-skills --skill pavlo-agent-ski…;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

pavlo-agent-skill-creator 怎么安装?

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

pavlo-agent-skill-creator 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。

pavlo-agent-skill-creator 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。

pavlo-agent-skill-creator 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。

pavlo-agent-skill-creator 用的人多吗?

公开数据是累计安装 267 次,近一周安装 11 次,下载 87 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。

pavlo-agent-skill-creator 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
pavlo-agent-skill-creator

巴甫洛特工技能创建器 · 本页

267—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

Skill: Agent Skill Creator

Create or edit SKILL.md files for AI coding agent skills in this repository.

File Location & Naming

  • Every skill lives in its own directory: .agents/skills/<skill-name>/SKILL.md
  • Skill directory names follow the pattern pavlo-<kebab-case-topic> (e.g., pavlo-jira-ticket-writer).
  • Ask for the skill name if the user did not specify one.

Examples Directory

Skills that produce output files (tickets, plans, PRs, etc.) should include an examples/ subdirectory with real-world reference files:

.github/skills/<skill-name>/
├── SKILL.md
└── examples/
    └── <real_example_file>.md
  • Place copies of real project files that demonstrate the skill's expected output.
  • The SKILL.md must reference the examples explicitly (e.g., "See examples/M2_PR3_plan.md for a real-world reference.").
  • Examples are snapshots — they do not need to stay in sync with the original files in docs/.
  • Agents should read the examples before generating output to match style and structure.

SKILL.md Structure

Every SKILL.md must follow this exact structure and section order.

1. YAML Metadata Block

A fenced YAML code block at the very top of the file — no content before it:

---
name: <skill-directory-name>
version: "1.0.0"
description: >
  <one or couple sentence description of what the skill does and when to use it.>
  Use when …
metadata:
  author: Pavlo Glushko
  applies_to:
    - "<glob pattern for files this skill produces or edits>"
  triggers:
    - <phrase that should activate this skill>
    - ...
  capabilities:
    - <specific thing the skill can do>
    - ...
---

Field rules:

  • name — must exactly match the directory name.
  • version — top-level document version; start at "1.0.0", increment minor for content additions, patch for corrections.
  • description — one or two sentences; include a "Use when …" clause so agents can decide whether to invoke it.
  • metadata.applies_to — glob(s) for the output or target files (e.g., "docs/ticket_*.md", "**/.agents/skills/**/SKILL.md").
  • metadata.triggers — lower-case phrases a user might say to invoke this skill. Include synonyms.
  • metadata.capabilities — concrete actions the skill can perform, not aspirational qualities.

2. Title

# Skill: <Human-Readable Skill Name>

One sentence immediately after the heading describing what the skill does (no heading needed for this sentence).

3. File Location & Naming *(if the skill produces output files)*

Explain where output files are created and the naming convention. Include concrete examples.

4. Body Sections

The remaining content is skill-specific. Structure it so an agent can follow it as a precise procedure. Common patterns used by existing skills:

  • Structure definition — when the skill produces a document, define every required section with its exact heading, content rules, and ordering (see pavlo-jira-ticket-writer).
  • Rules / conventions — when the skill enforces a coding or formatting standard, list rules grouped by category with explicit do/don't examples (see pavlo-markdown-editor).
  • Workflow — a numbered step-by-step procedure the agent follows when executing the skill.
  • Writing guidelines — tone, specificity level, privacy constraints, and anything else that cannot be inferred from the structure definition alone.

Workflow

Creating a new skill

  1. Determine the skill name: pavlo-<topic> in kebab-case.
  2. Create the file at .agents/skills/<skill-name>/SKILL.md.
  3. Write the YAML front matter block first (between --- delimiters).
  4. Add the # Skill: … title and one-sentence description.
  5. Add a File Location & Naming section if the skill produces output files.
  6. Write the body: structure definitions, rules, workflow, guidelines — whatever the skill needs to work without further clarification.
  7. Keep instructions agent-actionable: avoid vague advice; write steps an agent can execute directly.
  8. Add a row for the new skill to the skills table in .github/copilot-instructions.md — columns: Skill (display name), File (path to SKILL.md), Description (exact text from the description metadata field, collapsed to one line).

Editing an existing skill

  1. Read the existing SKILL.md in full before making any changes.
  2. Identify gaps: missing sections, outdated content, or conventions that changed in the codebase.
  3. Apply surgical updates — do not rewrite sections that are still accurate.
  4. Increment version in the metadata block (minor bump for additions, patch for corrections).
  5. If the description field changed, update the corresponding row in the skills table in .github/copilot-instructions.md to match exactly.
  6. Summarise what was added, modified, or removed so the user is aware of the changes.

Writing Guidelines

  • Be prescriptive, not descriptive. Write "use X" not "X is commonly used".
  • Include examples. Abstract rules are hard to follow; concrete examples eliminate ambiguity.
  • Reference real files. When a convention is grounded in an existing file, cite it by path.
  • One source of truth. If a convention is already in AGENTS.md or copilot-instructions.md, reference it rather than duplicating it.
  • Privacy. Do not expose personal workflow rules from copilot-instructions.md in skill output visible to others.

平台分布

Codex

35.55%
按下载量换算31

Claude

29.42%
按下载量换算26

Cursor

18.33%
按下载量换算16

Gemini CLI

9.43%
按下载量换算8

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