Mcp Server Commands

io.github.bytedancePublished by GitHub account bytedancev1.0.0Updated Oct 25, 2026

An MCP server to run arbitrary commands

VerifiedSTDIODesktop onlyCloud & InfrastructureDeveloper Tools

Overview

AI-generated overview

A local MCP server that lets an AI assistant execute arbitrary shell commands and run scripts on the user's machine.

What you can do
Run arbitrary commands on the local machine through the server's command execution tooling, with an optional working directory set at startup (for example --cwd). Execute script snippets by naming an interpreter and passing the script text, as shown in the in-memory call example using run_script with a node interpreter. Start the same server locally over stdio, or expose it as a Streamable HTTP or SSE endpoint on a chosen port for remote clients. Embed it in a JavaScript or TypeScript MCP client through in-process calls instead of installing the CLI.
When to use it
Useful when an assistant needs to run shell commands or short scripts on your own machine as part of a developer workflow, such as inspecting files, running build or test steps, or automating local tasks. It is a general-purpose execution bridge, so it fits setups where you already trust the client and want direct command access.
Requirements
Node.js 18 or newer and an MCP client such as VS Code, Cursor, Windsurf or Claude Desktop. Installed from the npm package @agent-infra/mcp-server-commands, typically launched with npx. No authentication, environment variables or headers are declared. Remote mode requires starting it with a port argument.
Before you install
It runs arbitrary commands on your machine, so anything the assistant is persuaded to execute runs with your user privileges and can read, change or delete files and reach the network. There is no declared authentication, so a remote HTTP or SSE endpoint should not be exposed beyond localhost. Review commands before approving them and consider restricting the working directory.

Installation

In SourceWeft

  1. Open Mcp Server Commands in the dashboard and add it to a workspace.
  2. Enable the server for the chats that should use its tools.

Desktop only via STDIO. STDIO servers start a local process, so they need the SourceWeft desktop host.

Other MCP clients

Follow the launch instructions in the repository.

README

Commands MCP Server

[NPM Downloads]

[Install MCP Server] [Install in VS Code] [Install in VS Code Insiders]

A Model Context Protocol (MCP) server that provides execuate arbitrary commands.

[Image]

Requirements

  • Node.js 18 or newer
  • VS Code, Cursor, Windsurf, Claude Desktop or any other MCP client

Getting started

Local (Stdio)

First, install the Commands MCP server with your client. A typical configuration looks like this:

js
{  "mcpServers": {    "commands": {      "command": "npx",      "args": [        "-y",        "@agent-infra/mcp-server-commands@latest",        "--cwd",        "/tmp"      ]    }  }}
Install in VS Code

You can also install the Commands MCP server using the VS Code CLI:

bash
# For VS Codecode --add-mcp '{"name":"commands","command":"npx","args":["@agent-infra/mcp-server-commands@latest"]}'

After installation, the Commands MCP server will be available for use with your GitHub Copilot agent in VS Code.

Install in Cursor

Go to Cursor Settings -> MCP -> Add new MCP Server. Name to your liking, use command type with the command npx @agent-infra/mcp-server-commands. You can also verify config or add command like arguments via clicking Edit.

js
{  "mcpServers": {    "commands": {      "command": "npx",      "args": [        "-y",        "@agent-infra/mcp-server-commands@latest"      ]    }  }}
Install in Windsurf

Follow Windsuff MCP documentation. Use following configuration:

js
{  "mcpServers": {    "commands": {      "command": "npx",      "args": [        "-y",        "@agent-infra/mcp-server-commands@latest"      ]    }  }}
Install in Claude Desktop

Follow the MCP install guide, use following configuration:

js
{  "mcpServers": {    "commands": {      "command": "npx",      "args": [        "-y",        "@agent-infra/mcp-server-commands@latest"      ]    }  }}
Remote (SSE / Streamable HTTP)

At the same time, use --port $your_port arg to start the browser mcp can be converted into SSE and Streamable HTTP Server.

bash
# normal run remote mcp servernpx @agent-infra/mcp-server-commands --port 8089

You can use one of the two MCP Server remote endpoint:

  • Streamable HTTP(Recommended): http://127.0.0.1:8089/mcp
  • SSE: http://127.0.0.1:8089/sse

And then in MCP client config, set the url to the SSE endpoint:

js
{  "mcpServers": {    "commands": {      "url": "http://127.0.0.1:8089/sse"    }  }}

url to the Streamable HTTP:

js
{  "mcpServers": {    "commands": {      "type": "streamable-http", // If there is MCP Client support      "url": "http://127.0.0.1:8089/mcp"    }  }}
In-memory call

If your MCP Client is developed based on JavaScript / TypeScript, you can directly use in-process calls to avoid requiring your users to install the command-line interface to use Commands MCP.

js
import { Client } from '@modelcontextprotocol/sdk/client/index.js';import { InMemoryTransport } from '@modelcontextprotocol/sdk/inMemory.js';
// type: module project usageimport { createServer } from '@agent-infra/mcp-server-commands';// commonjs project usage// const { createServer } = await import('@agent-infra/mcp-server-commands')
const client = new Client(  {    name: 'test commands client',    version: '1.0',  },  {    capabilities: {},  },);
const server = createServer();const [clientTransport, serverTransport] = InMemoryTransport.createLinkedPair();
await Promise.all([  client.connect(clientTransport),  server.connect(serverTransport),]);
// list toolsconst result = await client.listTools();console.log(result);
// call toolconst toolResult = await client.callTool({  name: 'run_script',  arguments: {    interpreter: 'node',    script: 'console.log(1+1);',  },});console.log(toolResult);

Developement

Access http://127.0.0.1:6274/:

bash
npm run dev

Source: packages/agent-infra/mcp-servers/commands/README.md at commit 2ff41a9

Tools

0
SourceWeft has not started this package in its sandbox yet, so its tools are not known.

Version history

1
  1. v1.0.0LatestSep 16, 2026

Score

50% of the applicable checks met

How complete this listing is, computed from the registry entry, the endpoint and public sources. Not a rating of the server and not a statement that it is safe or working today: reachability is a dated check, and no installs or reviews go into the score.

  • Reached in the last 30 days25 pts

    Keep the endpoint answering the MCP handshake; SourceWeft probes it every 30 days.

  • Auth declared10 pts

    Declare the credential headers in the registry entry, or keep the endpoint answering the probe so its auth can be classified.

  • Tools listed15 pts

    Answer tools/list without a credential, or ship a package that starts in a sandbox without one.

  • DescriptionRequired15 pts
  • Logo10 pts
  • Documentation15 pts
  • License10 pts