MCP of MCPs Server

Aggregates MCP servers to enable fast tool discovery, on-demand schemas, and low-token execution.

typescript

GitHub Stars

typescript

Language

6 months ago

First Indexed

2 months ago

Catalog Refreshed

Documentation & install

Readme and setup notes from the catalogue, plus a client-ready config you can copy for your MCP host.

Installation

Add the following to your MCP client configuration file.

Configuration

View docs

{
  "mcpServers": {
    "eliavamar-mcp-of-mcps": {
      "command": "npx",
      "args": [
        "-y",
        "@dbestai/mcp-of-mcps",
        "--config",
        "[{\"name\":\"weather\",\"command\":\"npx\",\"args\":[\"-y\",\"@h1deya/mcp-server-weather\"]},{\"name\":\"time\",\"command\":\"uvx\",\"args\":[\"mcp-server-time\"]}]"
      ]
    }
  }
}

MCP of MCPs is a meta-server that unifies multiple MCP servers into a single, smart endpoint. It enables fast tool discovery, on-demand loading of tool schemas, and efficient, low-token workflows by letting tools communicate directly. This helps you build scalable automation with less latency and fewer hallucinations, while preserving tool schemas across sessions for consistent behavior.

How to use

Start by listing available MCPs across your connected servers, then identify the tools you need by describing your intent in plain language. Load only the tool definitions you actually plan to use, and run your workflow so data flows directly between tools with only the final result returned to you. This pattern minimizes token usage and speeds up execution while reducing irrelevant information.

How to install

Prerequisites you need before installation:

Node.js is installed on your system
npm or yarn is available
You have MCP servers you want to aggregate

Install a local MCP of MCPs client or server package as part of your setup workflow, then configure the MCP connection as described in the configuration section.

Additional setup and configuration

{
  "mcpServers": {
    "mcp_of_mcps": {
      "autoApprove": [],
      "disabled": false,
      "timeout": 60,
      "type": "stdio",
      "command": "npx",
      "args": [
        "-y",
        "@dbestai/mcp-of-mcps",
        "--config",
        "[{\"name\":\"weather\",\"command\":\"npx\",\"args\":[\"-y\",\"@h1deya/mcp-server-weather\"]},{\"name\":\"time\",\"command\":\"uvx\",\"args\":[\"mcp-server-time\"]}]"
      ]
    }
  }
}

Additional setup and configuration (continued)

[
  {
    "name": "mcp_of_mcps",
    "command": "npx",
    "args": ["-y", "@dbestai/mcp-of-mcps", "--config",
      "[{\"name\":\"weather\",\"command\":\"npx\",\"args\":[\"-y\",\"@h1deya/mcp-server-weather\"]},{\"name\":\"time\",\"command\":\"uvx\",\"args\":[\"mcp-server-time\"]}]}"]
  },
  {
    "name": "mcp_of_mcps_file",
    "command": "npx",
    "args": ["-y", "@dbestai/mcp-of-mcps", "--config-file", "/absolute/path/to/your/config.json"]
  }
]

Available tools

semantic_search_tools

Semantic Discovery Tool that lets you search for tools by describing your task in plain English and returns the most relevant tools across all servers.

get_mcps_servers_overview

Discovery Tool that lists tool names only, providing a lightweight overview of available tools without loading full schemas.

get_tools_overview

Introspection Tool that loads full tool definitions for selected tools, including parameter schemas and usage examples.

run_functions_code

Execution Tool that enables data to flow directly between tools, performing the workflow in memory and returning only the final results.

Built by

VeilStrat

AI signals for GTM teams