Skip to main content
Glama
deenesjakoruzh
metatool-ai

MetaMCP MCP Server

by metatool-ai
OverviewSchemaSourceRelated ServersDiscussionsAdmin
TypeScript
Remote

MetaMCP MCP 服务器

https://metamcp.com ：一个 MCP 来管理您所有的 MCP

MetaMCP MCP 服务器是一个代理服务器，它将多个 MCP 服务器合并为一个。它从 MetaMCP 应用获取工具/提示/资源配置，并将工具/提示/资源请求路由到正确的底层服务器。

铁匠徽章

MetaMCP 应用程序仓库： https://github.com/metatool-ai/metatool-app

安装

通过 Smithery 安装

Smithery 有时可以正常工作（已在 Windsurf 本地确认），但有时不稳定，因为 MetaMCP 比较特殊，它在其上运行其他 MCP。如果无法正常工作，请考虑手动安装。

要通过Smithery自动为 Claude Desktop 安装 MetaMCP MCP 服务器：

npx -y @smithery/cli install @metatool-ai/mcp-server-metamcp --client claude

手动安装

export METAMCP_API_KEY=<env>
npx -y @metamcp/mcp-server-metamcp@latest
{
  "mcpServers": {
    "MetaMCP": {
      "command": "npx",
      "args": ["-y", "@metamcp/mcp-server-metamcp@latest"],
      "env": {
        "METAMCP_API_KEY": "<your api key>"
      }
    }
  }
}

Related MCP server: MCP of MCPs

用法

用作 stdio 服务器（默认）

mcp-server-metamcp --metamcp-api-key <your-api-key>

用作 SSE 服务器

mcp-server-metamcp --metamcp-api-key <your-api-key> --transport sse --port 12006

使用 SSE 传输选项，服务器将启动一个 Express.js Web 服务器，该服务器监听/sse端点上的 SSE 连接并接受/messages端点上的消息。

与 Docker 一起使用

在 Docker 容器内运行服务器并连接到主机上的服务时，使用--use-docker-host选项自动转换 localhost URL：

mcp-server-metamcp --metamcp-api-key <your-api-key> --transport sse --port 12006 --use-docker-host

这会将任何 localhost 或 127.0.0.1 URL 转换为host.docker.internal ，从而允许容器正确连接到主机上运行的服务。

命令行选项

Options:
  --metamcp-api-key <key>       API key for MetaMCP (can also be set via METAMCP_API_KEY env var)
  --metamcp-api-base-url <url>  Base URL for MetaMCP API (can also be set via METAMCP_API_BASE_URL env var)
  --report                      Fetch all MCPs, initialize clients, and report tools to MetaMCP API
  --transport <type>            Transport type to use (stdio or sse) (default: "stdio")
  --port <port>                 Port to use for SSE transport (default: "12006")
  --require-api-auth            Require API key in SSE URL path
  --use-docker-host             Transform localhost URLs to use host.docker.internal (can also be set via USE_DOCKER_HOST env var)
  -h, --help                    display help for command

环境变量

  • METAMCP_API_KEY ：MetaMCP 的 API 密钥

  • METAMCP_API_BASE_URL ：MetaMCP API 的基本 URL

  • USE_DOCKER_HOST ：设置为“true”时，将 localhost URL 转换为 host.docker.internal 以实现 Docker 兼容性

发展

# Install dependencies
npm install

# Build the application
npm run build

# Watch for changes
npm run watch

亮点

  • 与任何 MCP 客户端兼容

  • 多工作区层使您能够一键切换到另一组 MCP 配置。

  • MCP 配置的 GUI 动态更新。

  • 加入的 MCP 的命名空间隔离。

架构概述

sequenceDiagram
    participant MCPClient as MCP Client (e.g. Claude Desktop)
    participant MetaMCP-mcp-server as MetaMCP MCP Server
    participant MetaMCPApp as MetaMCP App
    participant MCPServers as Installed MCP Servers in Metatool App

    MCPClient ->> MetaMCP-mcp-server: Request list tools
    MetaMCP-mcp-server ->> MetaMCPApp: Get tools configuration & status
    MetaMCPApp ->> MetaMCP-mcp-server: Return tools configuration & status

    loop For each listed MCP Server
        MetaMCP-mcp-server ->> MCPServers: Request list_tools
        MCPServers ->> MetaMCP-mcp-server: Return list of tools
    end

    MetaMCP-mcp-server ->> MetaMCP-mcp-server: Aggregate tool lists
    MetaMCP-mcp-server ->> MCPClient: Return aggregated list of tools

    MCPClient ->> MetaMCP-mcp-server: Call tool
    MetaMCP-mcp-server ->> MCPServers: call_tool to target MCP Server
    MCPServers ->> MetaMCP-mcp-server: Return tool response
    MetaMCP-mcp-server ->> MCPClient: Return tool response

致谢

Deploy Server
One endpoint for every MCP toolRayrun runs every MCP server behind one URL and holds the credentials. Each call is decided per client, tool, and argument, then recorded.ray.runAd

Maintenance

ActivityInactive
ResponsivenessNo issues

Resources

Related MCP Connectors

Related MCP Servers

  • Magg

    sitbon
    A
    license
    B
    quality
    A
    maintenance
    A meta-MCP server that manages and aggregates other MCP servers, enabling LLMs to dynamically extend their own capabilities by searching for, adding, and configuring tool servers.
    16
    148 PyPI
    143
    AGPL 3.0

  • MCP of MCPs

    eliavamar
    A
    license
    Not graded
    quality
    D
    maintenance
    A meta-server that aggregates multiple MCP servers into a single interface, reducing token usage by 98%+ through progressive tool discovery and direct code execution that processes data between tools without consuming context window space.
    16 npm
    10
    Apache 2.0

  • MCPHub

    Jayden-Dong
    F
    license
    Not graded
    quality
    D
    maintenance
    A centralized management platform that aggregates multiple Model Context Protocol (MCP) servers into a single unified endpoint for AI agents. It provides a web interface for hot-swappable tool management, proxying of existing servers, and AI-powered generation of custom MCP plugins.
    1
    -

Appeared in Searches