SkillPorter

English | 中文

---

SkillPorter - 跨平台 Skill 转换工具

在不同 AI 编程助手之间无缝迁移你的技能配置。支持 Claude Code、Cursor、Cline、Kimi Code Agent 等 10+ 平台。

功能特性

• 🔄 一键转换 - 选择源平台和目标平台，即可完成 Skill 配置转换
• 👀 实时预览 - 转换前可预览结果，确认无误后再导出
• 📁 批量支持 - 支持单文件或整个 Skill 目录转换
• 🎯 智能识别 - 自动检测源文件格式，无需手动指定
• 🛡️ 变量映射 - 自动转换不同平台的变量语法（如 $ARGUMENTS ↔ {{args}}）
• 📦 一键导出 - 直接导出到目标平台目录

支持平台

平台	格式	路径
Claude Code	SKILL.md	~/.claude/skills/
OpenAI Codex	openai.yaml + AGENTS.md	~/.codex/skills/
WorkBuddy	SKILL.md	~/.workbuddy/skills/
CodeBuddy	SKILL.md	~/.codebuddy/skills/
Cursor	.cursorrules	项目根目录
Cline	.clinerules/*.md	项目根目录
KiloCode	.kilo/rules/*.md	项目根目录
Kimi Code Agent	SKILL.md	~/.kimi/skills/
通义灵码 CLI	SKILL.md	~/.qwen/skills/
Hermes	hermes.yaml	~/.hermes/skills/

快速开始

环境要求

• Python 3.9+
• PyQt6

安装

# 克隆仓库
git clone https://github.com/YOUR_USERNAME/skillporter.git
cd skillporter

# 安装依赖
pip install PyQt6 pyyaml

启动

python gui_launcher.py

使用说明

转换流程

1. 选择源平台 - 从下拉菜单选择源 Skill 的平台
2. 选择目标平台 - 选择要转换到的平台
3. 选择文件/文件夹 - 浏览或拖拽 Skill 文件到窗口
4. 预览转换 - 点击「预览转换」查看结果
5. 导出 - 点击「导出到目标目录」保存

转换模式

模式	说明
规则转换	快速、离线、免费，基于预设规则转换
API 智能转换	需联网，使用 LLM API 兜底处理复杂转换

项目结构

skillporter/
├── core/                    # 核心模块
│   ├── schema.py           # 通用 Skill 数据结构 (USS)
│   ├── parser.py           # 解析器基类
│   └── renderer.py         # 渲染器基类
├── platforms/               # 平台实现
│   ├── claude/             # Claude Code
│   ├── codex/              # OpenAI Codex
│   ├── workbuddy/          # WorkBuddy
│   ├── codebuddy/          # CodeBuddy
│   ├── cursor/             # Cursor
│   ├── cline/              # Cline
│   ├── kilocode/           # KiloCode
│   ├── kimi/               # Kimi Code Agent
│   ├── qwen/               # 通义灵码 CLI
│   └── hermes/             # Hermes
├── gui/                     # 图形界面
│   ├── main_window.py      # 主窗口
│   ├── convert_panel.py    # 转换面板
│   ├── preview_panel.py    # 预览面板
│   └── settings_dialog.py  # 设置对话框
├── config.py               # 配置管理
└── gui_launcher.py         # 启动入口

技术架构

• Universal Skill Schema (USS) - 跨平台通用数据结构，所有平台的 Skill 都会先转换为 USS 格式
• 解析器注册表 - 自动检测文件格式并选择合适的解析器
• 渲染器注册表 - 根据目标平台选择渲染器
• 变量转换引擎 - 自动映射不同平台的变量语法

开发

添加新平台

1. 在 core/schema.py 的 SkillPlatform 枚举中添加新平台
2. 在 platforms/ 下创建新目录，实现 parser.py 和 renderer.py
3. 更新 gui/convert_panel.py 的 PLATFORMS 字典

运行测试

python -m pytest tests/

版本历史

版本	说明
1.0	初始版本，支持 10 个平台转换

许可证

MIT License

---

SkillPorter - Cross-Platform Skill Conversion Tool

Seamlessly migrate your skill configurations between different AI programming assistants. Supports 10+ platforms including Claude Code, Cursor, Cline, and Kimi Code Agent.

Features

• 🔄 One-Click Conversion - Select source and target platforms, then convert instantly
• 👀 Real-time Preview - Preview results before exporting
• 📁 Batch Support - Convert single files or entire Skill directories
• 🎯 Smart Detection - Auto-detect source file format
• 🛡️ Variable Mapping - Auto-convert variable syntax across platforms (e.g., $ARGUMENTS ↔ {{args}})
• 📦 Quick Export - Export directly to target platform directory

Supported Platforms

Platform	Format	Path
Claude Code	SKILL.md	~/.claude/skills/
OpenAI Codex	openai.yaml + AGENTS.md	~/.codex/skills/
WorkBuddy	SKILL.md	~/.workbuddy/skills/
CodeBuddy	SKILL.md	~/.codebuddy/skills/
Cursor	.cursorrules	Project root
Cline	.clinerules/*.md	Project root
KiloCode	.kilo/rules/*.md	Project root
Kimi Code Agent	SKILL.md	~/.kimi/skills/
Tongyi Lingma CLI	SKILL.md	~/.qwen/skills/
Hermes	hermes.yaml	~/.hermes/skills/

Quick Start

Requirements

• Python 3.9+
• PyQt6

Installation

# Clone repository
git clone https://github.com/YOUR_USERNAME/skillporter.git
cd skillporter

# Install dependencies
pip install PyQt6 pyyaml

Launch

python gui_launcher.py

Usage

Conversion Workflow

1. Select Source Platform - Choose the source Skill's platform from dropdown
2. Select Target Platform - Choose the platform to convert to
3. Select File/Folder - Browse or drag-and-drop Skill files
4. Preview Conversion - Click "Preview Conversion" to see results
5. Export - Click "Export to Target Directory" to save

Conversion Modes

Mode	Description
Rule-based	Fast, offline, free conversion based on preset rules
API Smart	Online, uses LLM API as fallback for complex conversions

Project Structure

skillporter/
├── core/                    # Core modules
│   ├── schema.py           # Universal Skill Schema (USS)
│   ├── parser.py           # Parser base class
│   └── renderer.py         # Renderer base class
├── platforms/               # Platform implementations
│   ├── claude/             # Claude Code
│   ├── codex/              # OpenAI Codex
│   ├── workbuddy/          # WorkBuddy
│   ├── codebuddy/          # CodeBuddy
│   ├── cursor/             # Cursor
│   ├── cline/              # Cline
│   ├── kilocode/           # KiloCode
│   ├── kimi/               # Kimi Code Agent
│   ├── qwen/               # Tongyi Lingma CLI
│   └── hermes/             # Hermes
├── gui/                     # Graphical interface
│   ├── main_window.py      # Main window
│   ├── convert_panel.py    # Conversion panel
│   ├── preview_panel.py    # Preview panel
│   └── settings_dialog.py  # Settings dialog
├── config.py               # Configuration management
└── gui_launcher.py         # Entry point

Architecture

• Universal Skill Schema (USS) - Cross-platform universal data structure; all Skills are first converted to USS format
• Parser Registry - Auto-detect file format and select appropriate parser
• Renderer Registry - Select renderer based on target platform
• Variable Conversion Engine - Auto-map variable syntax across platforms

Development

Adding a New Platform

1. Add new platform to SkillPlatform enum in core/schema.py
2. Create new directory under platforms/ with parser.py and renderer.py
3. Update PLATFORMS dictionary in gui/convert_panel.py

Running Tests

python -m pytest tests/

Changelog

Version	Description
1.0	Initial release, supports 10 platforms

License

MIT License