skillsmcp
Exposes Agent Skills to AI agents as MCP tools, enabling discovery and activation of skill instructions for coding agents.
README
skillsmcp
MCP server that exposes Agent Skills to AI agents via the Model Context Protocol.
Agent Skills are reusable, portable instruction sets that guide AI coding agents. This server makes them discoverable and activatable as MCP tools, following the progressive disclosure pattern from the Agent Skills specification.
Prerequisites
- uv — fast Python package manager
Install with Homebrew:
brew install uv
Install
uv tool install git+https://github.com/aviddiviner/skillsmcp.git
This installs skillsmcp as a command on your PATH. To update later:
uv tool upgrade skillsmcp
Configure
Zed
Add to your settings (~/.config/zed/settings.json):
{
"context_servers": {
"skillsmcp": {
"command": "uvx",
"args": ["skillsmcp"]
}
}
}
Claude Desktop
Add to your Claude config (~/Library/Application Support/Claude/claude_desktop_config.json):
{
"mcpServers": {
"skillsmcp": {
"command": "uvx",
"args": ["skillsmcp"]
}
}
}
Other MCP Clients
Any client that supports MCP's stdio transport can use this server. Run it directly:
uvx skillsmcp
Tools
The server exposes three MCP tools:
| Tool | Description |
|---|---|
list_skills |
Discover all available skills with names and descriptions |
activate_skill |
Load a skill's full instructions by name |
read_skill_file |
Read supporting files from a skill's directory |
All tools accept an optional project_roots parameter — a list of project directories to scan for project-level skills. User-level skills (~/.agents/skills/, ~/.claude/skills/) are always included. When project_roots is not provided, the server falls back to its working directory.
Adding Skills
Skills are directories containing a SKILL.md file with YAML frontmatter:
~/.agents/skills/
└── my-skill/
├── SKILL.md
├── scripts/
│ └── helper.py
└── references/
└── REFERENCE.md
Example SKILL.md:
---
name: my-skill
description: A short description of what this skill does and when to use it.
---
# My Skill
Instructions for the AI agent go here...
Skill Directories
The server scans the following directories in precedence order (first-found wins for name collisions):
- Project-level (highest priority):
<project>/.agents/skills/<project>/.claude/skills/
- User-level:
~/.agents/skills/~/.claude/skills/
Project-level skills override user-level skills with the same name.
Development
git clone https://github.com/aviddiviner/skillsmcp.git
cd skillsmcp
# Install in development mode
uv sync
# Run the server directly
uv run skillsmcp
Learn More
- Agent Skills specification — the full format spec
- Example skills — official skill examples
- skills.sh — browse and install community skills
- FastMCP — the Python MCP framework powering this server
License
MIT
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。