VS Code MCP

v1.0.0Developer Toolsstable

MCP server for Claude Code/VSCode/Cursor/Windsurf to use editor self functionality. ⚡ Get real-time LSP diagnostics, type information, and code navigation for AI coding agents without waiting for slow tsc/eslint checks.

agentaiclaudecursoreslint
Share:
365
Stars
0
Downloads
0
Weekly
0/5

What is VS Code MCP?

VS Code MCP is a Model Context Protocol (MCP) server that allows AI assistants like Claude, Cursor, and VS Code to mcp server for claude code/vscode/cursor/windsurf to use editor self functionality. ⚡ get real-time lsp diagnostics, type information, and code navigation for ai coding agents without waiting for slow...

MCP server for Claude Code/VSCode/Cursor/Windsurf to use editor self functionality. ⚡ Get real-time LSP diagnostics, type information, and code navigation for AI coding agents without waiting for slow tsc/eslint checks.

This server falls under the Developer Tools category on MCPgee, the world's largest MCP server directory with 33,000+ servers.

Features

  • execute_shell_command
  • create_diff
  • open_file
  • open_project
  • check_extension_status

Use Cases

Real-time LSP diagnostics for AI
Code navigation and analysis
Type checking without slow tools
juehang

Maintainer

LicenseMIT License
Languagetypescript
Versionv1.0.0
UpdatedMay 19, 2026
Statushealthy
Maintenanceactive

Works with

ClaudeOpenAIwindowsmacoslinux

Installation

NPM

npx -y vscode

Manual Installation

npx -y vscode

Configuration

Configuration Details

Config File

claude_desktop_config.json

Performance

Response Metrics

Response Time< 200ms
ThroughputMedium

Resource Usage

Memory UsageLow
CPU UsageLow

How to Set Up and Use VS Code MCP

The VS Code MCP Server is a VS Code extension that exposes real-time editor capabilities — including LSP diagnostics, symbol search, file operations, and shell execution — to AI coding agents over the Model Context Protocol. Rather than waiting for slow standalone tsc or eslint runs, agents connected to this server can query live workspace errors and type information directly from the running language server. It is especially useful for Claude Code, Cursor, Windsurf, and similar AI coding environments where tight editor integration significantly reduces hallucinated fixes.

Prerequisites

  • VS Code installed with the vscode-mcp-server extension (marketplace ID: JuehangQin.vscode-mcp-server)
  • Node.js 18+ for running npx mcp-remote
  • An MCP-capable AI client such as Claude Desktop, Claude Code, or Cursor
  • The extension must be active in VS Code before the AI client connects
1

Install the VS Code extension

Open VS Code, go to the Extensions panel (Ctrl+Shift+X), search for 'vscode-mcp-server' by Juehang Qin, and install it. Alternatively install from the marketplace page at https://marketplace.visualstudio.com/items?itemName=JuehangQin.vscode-mcp-server.

2

Verify the extension is running

Open a workspace in VS Code. The extension starts an HTTP server on localhost:3000 by default. You can confirm by checking the VS Code status bar or Output panel for 'vscode-mcp-server started'. The port and host are configurable via VS Code settings: vscode-mcp-server.port and vscode-mcp-server.host.

3

Configure your MCP client to connect

Add the server to your MCP client configuration. The extension uses the streamable-HTTP transport via mcp-remote, so your client runs npx mcp-remote@next pointing at the extension's local server.

{
  "mcpServers": {
    "vscode-mcp-server": {
      "command": "npx",
      "args": ["mcp-remote@next", "http://localhost:3000/mcp"]
    }
  }
}
4

Adjust enabled tool categories (optional)

In VS Code Settings (Ctrl+,), search for 'vscode-mcp-server'. The 'enabledTools' setting lets you toggle individual categories: file (listing and reading), edit (create and modify), shell (terminal commands), diagnostics (errors and warnings), and symbol (definitions and outlines). Disable categories you do not want the AI agent to access.

5

Start using the server from your AI client

Begin a conversation in your AI client. The agent can now list files, read source, run diagnostics, search workspace symbols, and execute shell commands — all without leaving the conversation. A recommended workflow is to always begin with list_files_code on the root directory, then check diagnostics after each code change.

VS Code MCP Examples

Client configuration (Claude Desktop)

Add this block to your claude_desktop_config.json to connect Claude Desktop to the VS Code MCP server running on the default port.

{
  "mcpServers": {
    "vscode-mcp-server": {
      "command": "npx",
      "args": ["mcp-remote@next", "http://localhost:3000/mcp"]
    }
  }
}

Prompts to try

Example prompts that exercise the live LSP integration and file tools.

- "List all files in the root of my project and show me the directory structure"
- "Check my workspace for TypeScript errors and summarize them by file"
- "Find the definition of the UserService class and show me its public methods"
- "Run npm test in the terminal and report any failures"
- "Read src/api/client.ts and fix any type errors reported by the language server"

Troubleshooting VS Code MCP

Connection refused when the AI client tries to reach localhost:3000

Ensure the VS Code extension is installed and a workspace is open. The HTTP server only starts when the extension activates, which requires an open folder. Check the VS Code Output panel (select 'vscode-mcp-server' from the dropdown) for startup errors. If port 3000 is occupied by another process, change it via the vscode-mcp-server.port setting and update the URL in your MCP config accordingly.

Diagnostics tool returns no results even though errors exist in the editor

The extension reads diagnostics from VS Code's language server, which must have finished indexing the project. Wait a few seconds after opening the workspace for TypeScript/ESLint to complete their initial pass. Also verify the 'diagnostics' category is enabled in the vscode-mcp-server.enabledTools setting.

npx mcp-remote@next command not found or fails to install

Ensure Node.js 18 or later is installed and on your PATH. Run 'npm install -g mcp-remote' to pre-install the package globally, then update your config to use 'mcp-remote' directly instead of 'npx mcp-remote@next'.

Frequently Asked Questions about VS Code MCP

What is VS Code MCP?

VS Code MCP is a Model Context Protocol (MCP) server that mcp server for claude code/vscode/cursor/windsurf to use editor self functionality. ⚡ get real-time lsp diagnostics, type information, and code navigation for ai coding agents without waiting for slow tsc/eslint checks. It connects AI assistants to external tools and data sources through a standardized interface.

How do I install VS Code MCP?

Install via npm with the command: npx -y vscode. Then add the server configuration to your AI client's JSON config file (e.g., claude_desktop_config.json or .cursor/mcp.json).

Which AI clients work with VS Code MCP?

VS Code MCP works with all major MCP-compatible AI clients including Claude Desktop, Claude Code, Cursor, VS Code (GitHub Copilot), Windsurf, and Cline.

Is VS Code MCP free to use?

Yes, VS Code MCP is open source and available under the MIT License license. You can use it freely in both personal and commercial projects.

Browse More Developer Tools MCP Servers

Explore all developer tools servers available in the MCPgee directory. Each server includes setup guides for Claude, Cursor, and VS Code.

Quick Config Preview

{ "mcpServers": { "vscode": { "command": "npx", "args": ["-y", "vscode"] } } }

Add this to your claude_desktop_config.json or .cursor/mcp.json

Read the full setup guide →

Ready to use VS Code MCP?

Browse our complete directory of 33,000+ MCP servers, read setup guides for your editor, and start building with the Model Context Protocol.

33,000+ ServersFree & Open SourceStep-by-Step Guides