Desktop Commander MCP
@wonderwhy-er
Desktop Commander MCP について
This is MCP server for Claude that gives it terminal control, file system search and diff file editing capabilities
基本情報
設定
以下の設定を使って、このサーバーを MCP 対応クライアントに追加してください。
{
"mcpServers": {
"ClaudeDesktopCommander": {
"command": "npx",
"args": [
"@wonderwhy-er/desktop-commander@latest",
"remove"
]
}
}
}ツール
4`get_config`
`start_process`
`read_file`
`get_usage_stats`
概要
What is Desktop Commander MCP?
Desktop Commander MCP is a Model Context Protocol server that lets AI assistants (like Claude Desktop) search, update, manage files, and run terminal commands on your computer. It combines filesystem operations with interactive process control, supporting Excel, PDF, DOCX files, code execution in memory, and remote AI control from services like ChatGPT and Claude web. Built for developers and power users who want to manage their entire development workflow through an AI chat interface.
How to use Desktop Commander MCP?
Install via npx @wonderwhy-er/desktop-commander@latest setup (auto-updating), or use the bash installer, Smithery, manual config in claude_desktop_config.json, local git clone, or Docker. Most options provide automatic updates. After installation, restart Claude Desktop and use the provided tools in chat. For remote access, configure via Remote MCP at https://mcp.desktopcommander.app.
Key features of Desktop Commander MCP
- Remote AI control from ChatGPT, Claude web, and other services
- File preview UI with rendered markdown and inline images
- Interactive terminal commands with output streaming and timeout
- Execute Python, Node.js, and R code in memory without saving files
- Native support for Excel, PDF, and DOCX files (read, write, edit)
- Process management (list, kill, interact with running processes)
- Recursive directory listing with depth control and overflow protection
- Surgical text replacements, pattern-based search, and full file rewrites
- Audit logging with 10MB log rotation and detailed timestamps
- Security hardening: symlink prevention, command blocklist, Docker isolation
Use cases of Desktop Commander MCP
- Edit code and manage project files directly through AI chat
- Run long‑running commands (builds, servers, SSH sessions) with background execution
- Analyze CSV, JSON, or Excel data instantly by asking questions
- Interact with running processes (databases, dev servers) without leaving the chat
- Automate repetitive terminal tasks and file operations with AI assistance
FAQ from Desktop Commander MCP
How does Desktop Commander handle security?
The server includes symlink traversal prevention, a command blocklist with bypass protection, and optional Docker isolation for full sandboxing. See SECURITY.md for details.
Can I use Desktop Commander with other AI tools?
Yes—Desktop Commander supports Remote MCP, allowing use from ChatGPT, Claude web, and other AI services in addition to Claude Desktop.
Does Desktop Commander support Windows?
Yes. The README provides installation instructions for Windows via manual config, Docker (PowerShell script), and Smithery.
How do I update Desktop Commander?
Most installation methods (npx, bash script, Smithery, Docker) auto‑update when you restart Claude. For local git clones, run git pull && npm run setup manually.
What are the differences between Docker and native installation?
Docker runs in a sandboxed container with persistent volumes, requires no Node.js, and auto‑updates. Native installation gives direct filesystem access and uses Node.js. Both support folder mounting and the same tool set.
「生産性」の他のコンテンツ
MCPControl
CheffromspaceMCP server for Windows OS automation
Todo List MCP Server
RegiByteAn MCP server for managing todos within LLMs, created for educational purposes
applescript-mcp MCP Server
joshrutkowskiA macOS AppleScript MCP server
TickTick MCP Server
alexarevalo9A Model Context Protocol (MCP) server designed to integrate with the TickTick task management platform, enabling intelligent context-aware task operations and automation.
🚀 TaskMaster: Todoist MCP for Cursor AI
mingolladanieleA lightweight Model Context Protocol (MCP) server that enables natural language interaction with your Todoist tasks directly from your IDE. Built with simplicity and maintainability in mind.
コメント