wrapmcp
Universal MCP server that wraps any CLI tool, enabling AI assistants to run commands via natural language.
README
:suspect: Work/Test/Docs in Progress
wrapmcp
Wrap any CLI tool with MCP - Universal MCP server for well-documented CLI apps
Turn any command-line tool into an MCP server in seconds. No code required.
Why?
Modern CLI tools have excellent help systems and well-documented commands. wrapmcp makes them instantly available to AI assistants through the Model Context Protocol (MCP).
Instead of writing custom MCP servers for each tool, just wrap them:
wrapmcp install gh # GitHub CLI → MCP
wrapmcp install disco # Disco → MCP
wrapmcp install docker # Docker → MCP
wrapmcp install kubectl # Kubernetes → MCP
Installation
Download the latest binary for your platform from releases:
# Linux/macOS
curl -L https://github.com/yourusername/wrapmcp/releases/latest/download/wrapmcp-linux -o wrapmcp
chmod +x wrapmcp
# Install it
./wrapmcp install <your-favorite-cli-tool>
The binary will automatically copy itself to ~/.local/bin/wrapmcp and configure your MCP clients.
Usage
Install MCP server for a CLI tool
wrapmcp install disco
This will:
- Copy
wrapmcpto a stable location (~/.local/bin/) - Auto-detect your MCP clients (Claude Desktop, Claude Code, Continue, Zed)
- Add the tool to all detected clients
- You're done! Restart your AI assistant.
Target specific client
wrapmcp install gh --client=claude-desktop
wrapmcp install docker --client=claude-code
List installed servers
wrapmcp list
Uninstall
wrapmcp uninstall disco
Supported Clients
wrapmcp automatically detects and configures:
- Claude Desktop - Global installation
- Claude Code - Project-based (
.mcp.json) - Continue - VS Code extension
- Zed - Editor integration
How it works
When you run wrapmcp install disco, it creates an MCP server that:
- Registers a
run_clitool in your AI assistant - When called, executes
disco <args>and returns the output - The AI can explore commands using
--helpflags - Well-documented CLIs become instantly accessible
Examples
After installing, your AI assistant can do things like:
You: "List my disco projects"
AI: uses mcp__disco__run_cli with args "projects:list"
You: "Deploy my app to production"
AI: uses mcp__gh__run_cli to check status, then confirms with you
You: "Show running containers"
AI: uses mcp__docker__run_cli with args "ps"
Requirements
- The CLI tool must be installed and available in your
PATH - MCP client (Claude Desktop, Claude Code, Continue, or Zed)
Building from source
# Clone the repo
git clone https://github.com/yourusername/wrapmcp.git
cd wrapmcp
# Install dependencies
bun install
# Build
bun run build
# Use it
./wrapmcp install <tool>
License
MIT
Contributing
Issues and PRs welcome! This was built as a personal tool but happy to make it better.
wrapmcp - Because every great CLI deserves to be MCP-enabled ✨
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。