🛠️ LLM Skills Repository

English | 中文

Reusable skills and tools for LLMs - Python package publishing, FRP tunneling, and more automation scripts.

🔗 GitHub Repository: https://github.com/cicy-dev/llm-skills

---

English

📋 Available Skills

1. Python Package Publishing Skill

• Script: publish-pip-package.sh
• Documentation: PUBLISH_PIP_SKILL.md
• Function: Automated Python package publishing to PyPI
• Usage: ~/skills/publish-pip-package.sh [project_directory]

2. FRP Tunnel Setup Skill

• Script: setup-frp-tunnel.sh
• Documentation: SETUP_FRP_TUNNEL_SKILL.md
• Function: Setup FRP tunnels for SSH remote access
• Usage: ~/skills/setup-frp-tunnel.sh [server|client|colab] [parameters]

🚀 How to Use Skills

For LLMs:

# View all available skills
ls ~/skills/

# Use Python package publishing skill
~/skills/publish-pip-package.sh /path/to/python/project

# Use FRP tunnel skill
~/skills/setup-frp-tunnel.sh server
~/skills/setup-frp-tunnel.sh client SERVER_IP TOKEN

# View skill documentation
cat ~/skills/PUBLISH_PIP_SKILL.md

For Users:

1. Copy needed skill scripts to ~/skills/ directory
2. Ensure scripts have execute permissions: chmod +x ~/skills/script-name.sh
3. Create corresponding documentation files explaining usage

📁 Skills Directory Structure

~/skills/
├── README.md                    # This file
├── list-skills.sh              # Skills discovery tool
├── publish-pip-package.sh       # Python package publishing script
├── PUBLISH_PIP_SKILL.md        # Python package publishing documentation
├── setup-frp-tunnel.sh         # FRP tunnel setup script
├── SETUP_FRP_TUNNEL_SKILL.md   # FRP tunnel setup documentation
└── [other skill files...]

✨ Skill Features

• Reusable: Any LLM can invoke them
• Standardized: Unified interface and documentation format
• Self-contained: Each skill includes complete functionality and documentation
• Cross-project: Can be used across different projects

🔧 Adding New Skills

1. Create executable script file
2. Add corresponding .md documentation file
3. Update this README.md file
4. Ensure script has proper error handling and help information
5. Commit to GitHub: git add . && git commit -m "Add new skill" && git push

🌐 Install Skills Locally

# Clone skills repository
git clone https://github.com/cicy-dev/llm-skills.git ~/skills

# Or update existing skills
cd ~/skills && git pull

📖 Skill Naming Convention

• Script files: action-target.sh (e.g., publish-pip-package.sh)
• Documentation files: ACTION_TARGET_SKILL.md (e.g., PUBLISH_PIP_SKILL.md)
• Function description: Concise and clear, one sentence explanation

---

中文

📋 可用技能

1. Python 包发布技能

• 脚本: publish-pip-package.sh
• 文档: PUBLISH_PIP_SKILL.md
• 功能: 自动化发布 Python 包到 PyPI
• 使用: ~/skills/publish-pip-package.sh [项目目录]

2. FRP 隧道设置技能

• 脚本: setup-frp-tunnel.sh
• 文档: SETUP_FRP_TUNNEL_SKILL.md
• 功能: 设置 FRP 隧道实现 SSH 远程访问
• 使用: ~/skills/setup-frp-tunnel.sh [server|client|colab] [参数]

🚀 如何使用技能

对于 LLM:

# 查看所有可用技能
ls ~/skills/

# 使用 Python 包发布技能
~/skills/publish-pip-package.sh /path/to/python/project

# 使用 FRP 隧道技能
~/skills/setup-frp-tunnel.sh server
~/skills/setup-frp-tunnel.sh client SERVER_IP TOKEN

# 查看技能文档
cat ~/skills/PUBLISH_PIP_SKILL.md

对于用户:

1. 将需要的技能脚本复制到 ~/skills/ 目录
2. 确保脚本有执行权限: chmod +x ~/skills/script-name.sh
3. 创建对应的文档文件说明使用方法

📁 技能目录结构

~/skills/
├── README.md                    # 本文件
├── list-skills.sh              # 技能发现工具
├── publish-pip-package.sh       # Python 包发布脚本
├── PUBLISH_PIP_SKILL.md        # Python 包发布文档
├── setup-frp-tunnel.sh         # FRP 隧道设置脚本
├── SETUP_FRP_TUNNEL_SKILL.md   # FRP 隧道设置文档
└── [其他技能文件...]

✨ 技能特点

• 可重用: 任何 LLM 都可以调用
• 标准化: 统一的接口和文档格式
• 自包含: 每个技能都包含完整的功能和文档
• 跨项目: 可在不同项目中使用

🔧 添加新技能

1. 创建可执行脚本文件
2. 添加对应的 .md 文档文件
3. 更新本 README.md 文件
4. 确保脚本有适当的错误处理和帮助信息
5. 提交到 GitHub: git add . && git commit -m "Add new skill" && git push

🌐 安装技能到本地

# 克隆技能仓库
git clone https://github.com/cicy-dev/llm-skills.git ~/skills

# 或更新现有技能
cd ~/skills && git pull

📖 技能命名规范

• 脚本文件: action-target.sh (如: publish-pip-package.sh)
• 文档文件: ACTION_TARGET_SKILL.md (如: PUBLISH_PIP_SKILL.md)
• 功能描述: 简洁明了，一句话说明用途