home / mcp / local-command-server mcp server

local-command-server MCP Server

A TypeScript MCP server that executes shell commands and returns structured outputs for MCP clients.

Installation

Add the following to your MCP client configuration file.

Configuration

{
  "mcpServers": {
    "kentaro-mcp-local-command-server": {
      "command": "/path/to/mcp-local-command-server/build/index.js",
      "args": []
    }
  }
}

You have a TypeScript MCP server that runs shell commands and returns structured outputs to clients. It’s useful for automating tasks, integrating command execution into other tools, and querying command results in a consistent format.

How to use

You connect to this MCP server from an MCP client and request command executions. The server exposes a straightforward capability: you ask it to run a command, and it returns a structured result that you can parse and display in your workflow. Use this to automate routine tasks, fetch results from long-running commands, or integrate command execution into larger automation pipelines.

How to install

npm install

npm run build

npm run watch

To use with Claude Desktop, add the server config:

{
  "mcpServers": {
    "local-command-server": {
      "command": "/path/to/mcp-local-command-server/build/index.js"
    }
  }
}

Additional notes

Debugging is simplified by using the MCP Inspector, which you can run as part of development tooling. Start the inspector with:

npm run inspector

The inspector provides a URL to access debugging tools in your browser. This helps you inspect command execution, outputs, and any errors that occur during development.

Available tools

execute_command

Executes a shell command and returns a structured output containing stdout, stderr, exit code, and execution metadata.