# DesktopCommanderMCP

> MCP server that lets AI chat apps run terminal commands and edit files on your computer

- Repository: https://github.com/wonderwhy-er/DesktopCommanderMCP
- Category: Coding agents & dev tools
- Stars: 10,093 · Forks: 1,255 · Language: JavaScript · License: MIT
- Experience: Some experience
- Page: https://ir.adyog.com/repos/wonderwhy-er/desktopcommandermcp/

AI chat apps usually cannot touch your computer. Desktop Commander fixes that by running as an MCP (Model Context Protocol) server. The chat app connects to it and can then run commands, read and write files, and manage processes on your machine.

It builds on the official MCP Filesystem Server and adds search and edit tools. The AI can make small, targeted text replacements or rewrite whole files. It can search code with ripgrep and run code in memory (Python, Node.js, R) without saving a file. It also reads and edits Excel, PDF and Word files.

The README says it works through your client's subscription, so you do not pay API token costs. It works with Claude Desktop, Cursor, Windsurf, VS Code with Copilot, Cline and other MCP clients. The safety guardrails are a command blocklist and symlink protection. The README says plainly that these are not a sandbox. For stronger isolation it offers a Docker install.

## What it does

- **Terminal with process control**: Runs commands with streamed output, timeouts and background execution. It can talk to running processes such as SSH sessions, databases and dev servers, and list or kill processes.
- **File search and surgical edits**: Reads, writes, moves and searches files and folders, and uses ripgrep for code search. It can replace small pieces of text or rewrite a whole file.
- **Office and data file support**: Reads, writes and edits Excel (.xlsx, .xls, .xlsm), PDF and Word (.docx) files. You can ask it to analyze CSV, JSON or Excel data.
- **In-memory code execution**: Runs Python, Node.js or R code directly without saving a file first.
- **Local history and Docker option**: Tool calls are logged locally, with recent calls available through get_recent_tool_calls. A Docker install runs it in a container for isolation.

## What you could build

- Ask an AI assistant to search a codebase and make targeted edits across several files.
- Analyze a CSV, JSON or Excel file by asking questions in chat.
- Run and monitor long-running commands, such as a dev server or an SSH session, from a chat window.
- Create or edit Word documents and PDFs from markdown through an AI client.
- Run the server in Docker to give an AI limited access to chosen folders.

## Who it's for

Developers and power users of MCP-compatible AI clients who are comfortable with Node.js or Docker and understand that the AI will be able to run commands on their machine.

## Get started

With Node.js installed, run the setup command in a terminal, then restart Claude Desktop. For other clients, add the npx entry from the README to your MCP config file.

```
npx @wonderwhy-er/desktop-commander@latest setup
```

Tags: mcp, terminal, filesystem, claude-desktop, automation