Token导航 LogoToken导航TokenDH.com
研究检索敏感数据github未标认证来源可访问clear审计通过

claude-skill-managementClaude 技能 management

Agent Skill

claude-skill-management 用于查找、检索和筛选相关信息,适合在 Codex、Claude、Cursor、Gemini CLI 中需要根据关键词、任务场景或来源线索快速定位候选结果时使用。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。

总安装

1,011

周安装

43

GitHub Stars

12

下载量

354
CodexClaudeCursorGemini CLI

安装说明

本站只整理中文说明和来源信息,不托管安装包,也不代用户安装。

GitHub

来源数

3

许可证

MIT

最后核验

2026-05-01

来源状态

来源可访问

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:claude-skill-management(Claude 技能 management)
来源仓库:https://github.com/delphine-l/claude_global
仓库路径:skills/claude-skill-management
安装命令:
npx skills add https://github.com/delphine-l/claude_global --skill claude-skill-management
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。不同来源提供的安装方式可能略有差异;本站展示可直接复制的安装命令,安装前请核对来源页面。

skills.shnpx skills
npx skills add https://github.com/delphine-l/claude_global --skill claude-skill-management

简介

Claude Code 技能管理工具提供集中式仓库管理模式。

  • 支持 $CLAUDE_METADATA 元数据同步与符号链接项目关联机制。
  • 涵盖目录规范、命名约定、版本控制与团队协作等全流程管理指南。
  • 适合多人团队统一技能部署与更新,减少本地环境配置差异问题。
  • claude-skill-management 属于研究检索类 Skill,可作为该场景下的辅助能力补充。

SKILL.md

Claude Code Skill Management Expert

Expert knowledge for managing Claude Code skills and commands using the centralized repository pattern with $CLAUDE_METADATA.

Supporting Documentation

This skill is split across multiple files for maintainability. Read these as needed:

When to Use This Skill

  • Creating new global skills or commands
  • Setting up skills for a new project
  • Synchronizing projects with updated global skills
  • Organizing the centralized skill repository
  • Troubleshooting skill discovery or activation issues
  • Understanding the skill lifecycle

Environment Setup

Required Environment Variable

$CLAUDE_METADATA must be set to your centralized skills directory.

Check if set:

echo $CLAUDE_METADATA
# Should output your claude_data directory path

If not set, add to ~/.zshrc (or ~/.bashrc):

export CLAUDE_METADATA="$HOME/path/to/claude_data"  # Adjust to your actual path

Apply immediately:

source ~/.zshrc  # or source ~/.bashrc

Verify Directory Structure

ls -la $CLAUDE_METADATA/
# Should show:
# ├── skills/       # Global skills
# ├── commands/     # Global commands
# ├── hooks/        # Claude Code hooks (symlinked to ~/.claude/hooks/)
# ├── README.md
# └── QUICK_REFERENCE.md

Complete Setup from Scratch

If setting up a centralized skill repository for the first time:

  1. Create directory structure: mkdir -p $CLAUDE_METADATA/{skills,commands} cd $CLAUDE_METADATA
  2. Set environment variable (add to ~/.zshrc or ~/.bashrc): echo 'export CLAUDE_METADATA="$HOME/path/to/claude_data" # Adjust to your actual path' >> ~/.zshrc source ~/.zshrc
  3. Verify setup: echo $CLAUDE_METADATA # Should output your claude_data directory path
  4. Create initial documentation: # Create README and QUICK_REFERENCE # (use templates from claude-skill-management skill)
  5. Initialize git (recommended): cd $CLAUDE_METADATA git init git add. git commit -m "Initial centralized skill repository"
  6. Create your first skill: mkdir -p $CLAUDE_METADATA/skills/my-first-skill # Create SKILL.md with frontmatter
  7. Link to first project: cd ~/Workdir/my-project mkdir -p.claude/skills ln -s $CLAUDE_METADATA/skills/my-first-skill.claude/skills/

Environment variable best practices:

  • Use $HOME not hardcoded paths for portability
  • Source shell config after adding: source ~/.zshrc
  • Verify in new terminals: echo $CLAUDE_METADATA
  • Document for team members in README.md

Creating New Skills

Step 1: Create Skill Directory

mkdir -p $CLAUDE_METADATA/skills/your-skill-name

Naming conventions:

  • Use kebab-case (lowercase with hyphens)
  • Be descriptive but concise
  • Examples: galaxy-tool-wrapping, python-testing, docker-workflows

Step 2: Create SKILL.md with Frontmatter

cat > $CLAUDE_METADATA/skills/your-skill-name/SKILL.md << 'EOF'
---
name: your-skill-name
description: Brief description that helps Claude decide when to activate this skill (1-2 sentences)
---

# Your Skill Name

Detailed instructions for Claude when this skill is activated.

## When to Use This Skill

- Specific use case 1
- Specific use case 2
- Specific use case 3

## Core Concepts

### Concept 1

Explanation and examples...

### Concept 2

Explanation and examples...

## Best Practices

- Practice 1
- Practice 2

## Common Issues and Solutions

### Issue 1

**Problem:** Description
**Solution:** How to fix it

## Examples

### Example 1: Task Name

Description and code examples...
EOF

Frontmatter fields:

  • name (required): Must match directory name
  • description (required): Clear, concise description for activation
  • version (optional): Semantic versioning (e.g., 1.0.0)
  • dependencies (optional): Required tools/packages

Step 3: Add Supporting Files (Optional)

# Add detailed reference documentation
cat > $CLAUDE_METADATA/skills/your-skill-name/reference.md << 'EOF'
# Reference Documentation

Detailed technical information, API references, etc.
EOF

# Add examples directory
mkdir -p $CLAUDE_METADATA/skills/your-skill-name/examples

# Add templates directory
mkdir -p $CLAUDE_METADATA/skills/your-skill-name/templates

Step 4: Test the Skill

# Create a test project
mkdir -p /tmp/test-skill-project/.claude/skills

# Symlink the new skill
ln -s $CLAUDE_METADATA/skills/your-skill-name /tmp/test-skill-project/.claude/skills/your-skill-name

# Start Claude Code in test project
cd /tmp/test-skill-project
# Tell Claude: "Use the your-skill-name skill to [test task]"

Creating New Commands

Step 1: Choose or Create Category Directory

# Use existing category
ls $CLAUDE_METADATA/commands/
# Or create new category
mkdir -p $CLAUDE_METADATA/commands/your-category

Common categories:

  • vgp-pipeline/ - VGP workflow commands
  • git-workflows/ - Git-related commands
  • testing/ - Testing-related commands
  • deployment/ - Deployment commands

Step 2: Create Command File

cat > $CLAUDE_METADATA/commands/your-category/command-name.md << 'EOF'
---
name: command-name
description: Brief description shown in /help
---

Your command prompt here. This will be expanded when the user types /command-name.

You can include:
- Multi-line instructions
- Variable references: {{variable_name}}
- Markdown formatting
- Code blocks

Example:
Check the status of all workflows for species {{species_name}}.
Show me which workflows are complete, running, or failed.
EOF

Naming conventions:

  • Use kebab-case
  • Start with verb: check-status, debug-failed, update-skills
  • Be specific: deploy-production not just deploy

Step 3: Test the Command

# Symlink to test project
ln -s $CLAUDE_METADATA/commands/your-category/command-name.md /tmp/test-project/.claude/commands/

# Start Claude Code and test
# Type: /command-name

Command Help System

Viewing Command Documentation

Use /command-help to view documentation for Claude Code commands (similar to --help in traditional CLI tools):

# List all available commands
/command-help list

# Show specific command help
/command-help share-project

# Show full details including implementation steps
/command-help share-project --full

Command Help Implementation

Location: $CLAUDE_METADATA/commands/global/command-help.md

Features:

  • Lists global and project commands with descriptions
  • Shows usage, parameters, and examples
  • Can display full implementation steps with --full flag
  • Searches in both global and project command directories

Command Frontmatter Format

Commands should include frontmatter for the help system:

---
description: Brief one-line description
usage: /command-name [arguments]
parameters: |
  arg1: Description of argument 1
  arg2: Description of argument 2
examples: |
  /command-name example1
  /command-name example2 --option
---

[Command implementation steps...]

Creating Help-Enabled Commands

Template for new commands:

---
name: my-command
description: Brief description of what this command does
usage: /my-command [required-arg] [optional-arg]
parameters: |
  required-arg: Description of required argument
  optional-arg: (Optional) Description of optional argument
examples: |
  /my-command basic-example
  /my-command advanced-example --flag
---

# Command Implementation

Step 1: [First step description]

Step 2: [Second step description]

[Continue with detailed steps...]

Best practices for command documentation:

  1. Keep description to 1 line (shows in list view)
  2. Document all parameters clearly
  3. Provide realistic examples
  4. Include expected output in steps
  5. Note any prerequisites or dependencies

For additional details, see the supporting files listed at the top of this document:

  • Symlinking and project setup: symlinking-guide.md
  • Updating and syncing: updating-and-syncing.md
  • Repository organization: repository-organization.md
  • Version control and git: version-control.md
  • Troubleshooting: troubleshooting.md
  • Best practices: best-practices.md
  • Quick reference and workflows: quick-reference.md

适合场景

01

用户想查找某类 Agent Skill 时

02

需要根据任务场景推荐可安装能力包时

03

需要对比不同来源的安装命令和来源信息时

04

需要参考平台分布和安装热度时

能力概览

能力 1

按任务关键词查找相关 Skills

能力 2

展示可复制的安装命令

能力 3

保留来源站点、仓库和原始说明,方便继续核验

能力 4

补充不同宿主或平台的使用分布数据

能力 5

展示第三方安全扫描或审计结果

安装后应在对应宿主中按原始 README 的触发条件使用;具体调用方式请以来源页面和 README 为准。

平台分布

Claude Code

30.34%
按下载量换算107

windsurf

21.88%
按下载量换算77

OpenCode

16.3%
按下载量换算58

Codex

11.2%
按下载量换算40

Antigravity

8.79%
按下载量换算31

Gemini CLI

3.48%
按下载量换算12

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

通过

权限和风险

敏感数据

该 Skill 可能接触密钥、Token、环境变量或敏感配置,应进入高风险复核队列,默认不自动发布。

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。

来源信息

继续浏览同类 Skills