通过渐进披露，将Claude的技能系统带入任何兼容MCP的AI代理。通用、基于标准且易于使用。
概览
一个通用MCP服务器，使支持MCP的代理应用能够使用Anthropic官方的Claude代理技能，同时降低上下文开销并最大化能力。 Claude Agent Skills设计精妙，但仅限于Claude的生态系统。本项目打破了这一限制，包括： - ✅ ： 支持任何兼容MCP的代理（Kilo Code、Cursor、Roo Code、Codex等）- ✅ ： 使用官方Anthropic Skill格式 - 无需修改 - ✅ ： 实现与Claude Code相同的智能上下文加载 - ✅ ： 标准MCP协议意味着你永远不受限于单一平台 传统的MCP工具会在你开始前就加载所有文档，导致你在开始前就消耗大量令牌。有了15+工具，你的代理在做任何实际工作之前就已经缺乏上下文。 通过渐进披露：代理最初看到轻量级技能列表，然后仅在需要时加载完整细节。该项目为每个兼容MCP的代理带来了同样的效率。
概述（AI 翻译）
将Claude特工技能带到任何兼容MCP的特工上
一个通用MCP服务器，使支持MCP的任何代理应用都能使用Anthropic官方的Claude代理技能，并实现渐进披露——减少上下文开销同时最大化能力。
包名：agentskill-mcp |PyPI：agentskill-mcp
英文 |简体中文
为什么是AgentSkill MCP？
Claude Agent Skills设计精妙，但仅限于Claude的生态系统。本项目打破了这一限制，包括：
- ✅ 通用兼容性：支持任何兼容MCP的代理（Kilo Code、光标、Roo Code、Codex等）
- ✅ 100% Claude技能兼容：使用官方拟人技能格式 - 无需修改
- ✅ 渐进披露：实现与Claude代码相同的智能上下文加载
- ✅ 零锁定：标准MCP协议意味着你永远不会被绑定在一个平台上
技能解决问题
传统的MCP工具会在你开始前就加载所有文档，导致你在开始前就消耗大量令牌。有了15+工具，你的代理在做任何实际工作之前就已经缺乏上下文。
本页由 AI 翻译，技术细节请以原文为准。
README
以下为原文 README 节选，完整内容请查看原文。
AgentSkill MCP
Bring Claude Agent Skills to ANY MCP-compatible Agent
A universal MCP server that enables any Agent application with MCP support to use Anthropic’s official Claude Agent Skills with progressive disclosure - reducing context overhead while maximizing capability.
Package Name: agentskill-mcp | PyPI: agentskill-mcp
English | 简体中文
Why AgentSkill MCP?
Claude Agent Skills are brilliantly designed but locked to Claude’s ecosystem. This project breaks that limitation by:
- ✅ Universal Compatibility: Works with ANY MCP-compatible agent (Kilo Code, Cursor, Roo Code, Codex, and more)
- ✅ 100% Claude Skill Compatible: Uses official Anthropic Skill format - no modifications needed
- ✅ Progressive Disclosure: Implements the same smart context loading as Claude Code
- ✅ Zero Lock-in: Standard MCP protocol means you’re never tied to one platform
The Problem Skills Solve
Traditional MCP tools load ALL documentation upfront, consuming massive amounts of tokens before you even start. With 15+ tools, your agent is context-starved before doing any real work.
Skills fix this through progressive disclosure: agents see a lightweight skill list initially, then load full details only when needed. This project brings that same efficiency to every MCP-compatible agent.
Features
- 🚀 One-Line Installation:
pip install agentskill-mcporuvx agentskill-mcp - 🔌 Universal MCP Compatibility (gradual testing): Works with Kilo Code, Cursor, Roo Code, Codex, Cherry Studio, and any MCP-compatible agent
- 📦 Official Skill Format: Fully compatible with Anthropic’s Claude Skills
- 🎯 Progressive Disclosure: Smart context loading - minimal overhead until skills are needed
- 🔄 Hot Reload (not yet implemented): File changes detected and updated in real-time (where protocol supported)
- 🗂️ Smart Path Discovery: Auto-detects
.claude/skills/,.skill/, or custom directories - 🌍 Environment Aware: Project-level and global skill directories with automatic detection
- 🎨 ClaudeCode Compatible: Supports both
.claude/skills/(ClaudeCode format) and.skill/(custom format for this project)
Project Status
⚠️ Early Development - This project is in early stages. Currently tested on Windows only.
Tested Platforms:
- ✅ Kilo Code (AI coding assistant) - Windows
- ✅ Roo Code (AI coding assistant) - Windows
- ✅ Cline (AI coding assistant) - Windows
Next Steps:
- 🔄 Testing on more MCP-compatible agents (Codex, Cursor, QwenCode, etc.)
- 🔄 Cross-platform testing (macOS, Linux)
- 🔄 Broader compatibility verification
- 🔄 Hot reload feasibility testing (based on MCP’s List Changed Notification) see: https://modelcontextprotocol.io/specification/2025-06-18/server/tools#list-changed-notification
安装
uvx agentskill-mcp --skills-dir C:\\Users\\YourName\\path\\to\\skills配置
{
"mcpServers": {
"skills": {
"command": "uvx",
"args": [
"agentskill-mcp",
"--skills-dir",
"C:\\Users\\YourName\\path\\to\\skills"
]
}
}
}