Client Guide

Best MCP Servers for VS Code / GitHub Copilot (2026)

Visual Studio Code with GitHub Copilot supports MCP servers for extending AI capabilities. Configure servers in VS Code settings to connect Copilot to databases, APIs, and local tools.

60+ Compatible Servers
JSON Config

What is VS Code with GitHub Copilot MCP?

Visual Studio Code is the world's most popular code editor, and with the GitHub Copilot extension, it now supports the Model Context Protocol (MCP). This integration allows Copilot Chat to interact with external tools, file systems, databases, GitHub repositories, and custom MCP servers - giving the AI deep context about your project beyond just the files open in your editor.

MCP support in VS Code is configured through the standard settings.json file under the github.copilot.chat.mcp.servers key. This integrates naturally with VS Code's existing settings system, making it familiar for anyone who has customized VS Code before.

How MCP Works with VS Code

When you configure MCP servers in VS Code's settings, the GitHub Copilot extension manages the server lifecycle. It starts configured servers when you open a project, maintains connections, and exposes the servers' tools within Copilot Chat. You can then ask Copilot questions that require external data - like querying a database, searching the web, or reading files from specific directories - and Copilot will use the appropriate MCP tools to get the information it needs.

The integration works within Copilot Chat (the sidebar chat panel). When MCP tools are available, Copilot can autonomously decide when to use them based on your questions, or you can explicitly reference tools in your prompts.

Detailed Setup Process

  1. Install VS Code from code.visualstudio.com if you do not already have it.
  2. Install the GitHub Copilot extension from the VS Code marketplace. You need an active GitHub Copilot subscription (Individual, Business, or Enterprise).
  3. Open your project in VS Code.
  4. Open project settings: Create or edit .vscode/settings.json in your project root. For global settings, use VS Code's user settings (Cmd+Shift+P > "Preferences: Open User Settings (JSON)").
  5. Add MCP server configurations under the github.copilot.chat.mcp.servers key. Note that this key is different from other clients - it uses the Copilot-specific namespace.
  6. Save the settings file. VS Code may prompt you to start the configured servers. Accept the prompt to activate them.
  7. Open Copilot Chat (Cmd+Shift+I on macOS, Ctrl+Shift+I on Windows/Linux) and verify that MCP tools appear in the available tools list.
  8. Start using Copilot with your MCP servers - ask questions that require external data and Copilot will use your configured tools.

Advanced Configuration

Workspace vs. User Settings

VS Code has two levels of settings: User settings (global, apply to all projects) and Workspace settings (project-specific, stored in .vscode/settings.json). You can define MCP servers at either level. Workspace settings take precedence, allowing you to have project-specific server setups while maintaining a global baseline. This is similar to how Cursor handles project vs. global configuration.

Environment Variables

The env field in each server configuration lets you pass environment variables like API keys and database credentials. For team projects, use workspace settings that reference variable names and have each developer set the values in their system environment or in a .env file. This way, the config can be committed to version control without exposing secrets.

Transport Options

VS Code supports stdio transport (using command and args) and SSE transport (using a url field). Stdio is the standard for local servers. SSE is useful for remote servers, shared team services, or servers that need to handle multiple concurrent connections.

Integration with Other VS Code Extensions

MCP servers in VS Code complement other extensions. For example, you can use the PostgreSQL MCP server alongside a database visualization extension, or the GitHub MCP server alongside the GitHub Pull Requests extension. The MCP tools give Copilot additional context that makes its suggestions more accurate and relevant.

Troubleshooting Common Issues

  • "MCP tools not showing in Copilot Chat" - Ensure you have the latest version of the GitHub Copilot extension. MCP support was added in a specific version, so an outdated extension will not show MCP features. Also verify that your Copilot subscription is active.
  • "Server failed to start" - Check the Output panel (View > Output) and select "GitHub Copilot" from the dropdown. Look for MCP-related error messages. Common causes: the command is not in PATH, missing Node.js installation, or incorrect arguments.
  • "Permission denied when starting server" - On macOS/Linux, ensure the server command has execute permissions. For npx-based servers, verify that your npm/node installation is accessible from VS Code's environment (which may differ from your terminal's environment).
  • Config key not recognized - Make sure you use github.copilot.chat.mcp.servers (not mcpServers like other clients). This is the Copilot-specific namespace in VS Code settings.
  • Servers work in terminal but not in VS Code - VS Code may use a different shell environment than your terminal. Check VS Code's terminal.integrated.env.* settings to ensure necessary environment variables and PATH entries are available.

Recommended MCP Servers

These MCP servers integrate particularly well with the VS Code development workflow:

  • Filesystem Server - Access files outside your workspace, or provide sandboxed file access to specific directories.
  • GitHub Server - Manage issues, pull requests, and code reviews directly from Copilot Chat.
  • PostgreSQL Server - Query databases while coding, explore schemas, and generate SQL queries with natural language.
  • Brave Search Server - Search documentation, Stack Overflow, and other resources without leaving the editor.
  • Memory Server - Store project context and decisions that persist across VS Code sessions.

Browse our full MCP server directory for all compatible servers.

Related Guides

For a side-by-side comparison of MCP setup in VS Code, Cursor, and Claude, read our MCP Servers for Cursor, VS Code, and Claude guide. Our tutorials section has step-by-step walkthroughs for common server configurations.

Performance Tips

  • Configure only what you need. Each MCP server is a separate process consuming memory and CPU. Limit your configuration to servers you actively use.
  • Use workspace settings to load project-specific servers instead of having everything in user settings. This keeps each project lean.
  • Pre-install frequently used packages globally with npm install -g to avoid the npx -y download step on each startup.
  • Monitor resource usage through VS Code's Process Explorer (Help > Open Process Explorer) to identify servers consuming excessive resources.
  • Disable unused servers by commenting them out in settings.json rather than deleting them, so you can easily re-enable them later.

Security Considerations

MCP servers run with your user permissions on your local machine. Follow these practices for safe usage:

  • Protect credentials. Never commit API keys or tokens in .vscode/settings.json to version control. Use environment variables or add the file to .gitignore.
  • Limit filesystem scope. When using the filesystem server, restrict it to specific project directories rather than giving it access to your entire home directory or system root.
  • Verify server sources. Only install MCP servers from trusted repositories. Review the source code of any server that accesses sensitive data.
  • Review tool calls. Copilot shows what MCP tools it wants to use. Review these requests before approving, especially for servers that can modify files, databases, or make API calls.
  • Keep Copilot updated. GitHub regularly releases security patches and improvements for the Copilot extension, including its MCP integration.
  • For comprehensive security guidance, read our MCP Server Security Guide.

Comparing VS Code with Other Clients

VS Code with Copilot is the best choice if you are already invested in the VS Code ecosystem and have a Copilot subscription. For a more AI-first editor built from the ground up, consider Cursor. For an autonomous agent experience within VS Code, try Cline. For a standalone chat interface, use Claude Desktop. All these clients support the same MCP servers - the only difference is the configuration format and key names.

Configuration

Config file location: .vscode/settings.json

{ "github.copilot.chat.mcp.servers": { "filesystem": { "command": "npx", "args": ["-y", "@modelcontextprotocol/server-filesystem", "."] }, "github": { "command": "npx", "args": ["-y", "@modelcontextprotocol/server-github"], "env": { "GITHUB_PERSONAL_ACCESS_TOKEN": "ghp_your_token_here" } }, "postgres": { "command": "npx", "args": ["-y", "@modelcontextprotocol/server-postgres", "postgresql://localhost:5432/mydb"] }, "brave-search": { "command": "npx", "args": ["-y", "@modelcontextprotocol/server-brave-search"], "env": { "BRAVE_API_KEY": "your_brave_api_key_here" } }, "memory": { "command": "npx", "args": ["-y", "@modelcontextprotocol/server-memory"] } } }

Setup Guide

1

Install VS Code from code.visualstudio.com and the GitHub Copilot extension from the marketplace.

2

Ensure you have an active GitHub Copilot subscription (Individual, Business, or Enterprise).

3

Open your project and create or edit .vscode/settings.json.

4

Add MCP server configurations under the github.copilot.chat.mcp.servers key.

5

Save the settings file. VS Code may prompt you to start the configured servers.

6

Open Copilot Chat (Cmd+Shift+I / Ctrl+Shift+I) and verify MCP tools appear in the available tools list.

7

Use Copilot Chat with your MCP servers - Copilot will use them to answer questions and perform actions.

Need help setting up VS Code / GitHub Copilot?

Check our step-by-step IDE setup guide with troubleshooting tips.

Read Setup Guide

Explore VS Code / GitHub Copilot Servers by Category

Find the best MCP servers for VS Code / GitHub Copilot in each category.

File Systems

MCP servers for secure file operations, directory management, and document processing. These servers provide sandboxed access to local and remote file systems with configurable permissions.

Databases

MCP servers for connecting AI assistants to SQL and NoSQL databases. Query, analyze, and manage your data through natural language with support for PostgreSQL, SQLite, MongoDB, Redis, and more.

APIs

MCP servers that connect AI assistants to external APIs and web services. Search the web, fetch data, interact with third-party platforms, and automate API workflows through natural language.

Cloud Services

MCP servers for managing cloud infrastructure across AWS, Google Cloud, Azure, and platforms like Vercel, Netlify, and Cloudflare. Deploy, monitor, and manage cloud resources through AI assistants.

Developer Tools

MCP servers for software development workflows including version control, CI/CD, code analysis, browser testing, and project management. Supercharge your development process with AI-powered tooling.

Analytics

MCP servers for monitoring, observability, and data analytics. Connect AI assistants to Grafana, Datadog, and search platforms to analyze metrics, logs, and business data in real time.

Communication

MCP servers for messaging, video conferencing, and team collaboration platforms. Connect AI assistants to Slack, Discord, and Zoom for automated communication workflows.

Business Applications

MCP servers for CRM, e-commerce, project management, and business automation platforms. Connect AI to Shopify, Stripe, Salesforce, HubSpot, Notion, and more to streamline business operations.

Frequently Asked Questions

Ready to supercharge VS Code / GitHub Copilot with MCP?

Browse our complete directory of 60+ MCP servers, read our setup guides, and start building with the Model Context Protocol today.

60+ ServersFree & Open SourceStep-by-Step GuidesSecurity Reviews