Claude MCP Servers: What They Are and How to Add One

MCP is how Claude connects to tools and data outside itself. Adding a server is one terminal command.

Last updated August 2026

Heads up: some links below are referral links. They cost you nothing extra and never change what we recommend. How this site makes money.

The short answer

MCP (Model Context Protocol) is an open standard for connecting Claude to external tools and data. An MCP server is a small program exposing a capability — a database, an API, your filesystem — that Claude can then use.

Adding one in Claude Code is a single command:
claude mcp add --transport http <name> <url>

What MCP actually solves

A language model on its own can only reason about what's in its context. MCP is the standard plumbing for giving it access to things outside that — your issue tracker, a database, a documentation site, your local files.

The reason it matters is that it's a standard. Before it, every integration was bespoke. Now a server written for one MCP-compatible client works with others, and there's a growing ecosystem to draw on rather than building each connection yourself.

Adding a remote server

Run this in your terminal, not inside a Claude session — you're configuring the server before starting a conversation:

claude mcp add --transport http claude-code-docs https://code.claude.com/docs/mcp

The name is yours to choose; the URL comes from whoever publishes the server.

Configuration scopes

Use --scope user for anything you want everywhere, --scope project for something the whole team needs.

Local servers

A local stdio server is a program Claude Code launches as a subprocess on your machine rather than reaching over a URL.

Use one when the tool needs local resources — your filesystem, a database socket, a browser. Since it runs on your machine it can reach things no remote server could.

Adding servers outside the terminal

You don't have to use the CLI. In the Claude desktop app, go to Settings → Extensions, click Browse extensions for the directory, and install what you want.

Servers can also be added from VS Code and the web. The desktop directory is the easiest starting point if you'd rather not touch a terminal.

The security part

An MCP server is code you're granting access to your data and tools. Install servers the way you'd install any dependency: from sources you trust, and understanding what access you're handing over.

Be especially careful with servers that can both read sensitive data and reach the network — that combination is what makes data exfiltration possible if a server turns out to be malicious or compromised.

Common questions

What does MCP stand for?
Model Context Protocol — an open standard for connecting AI models to external tools and data.
Do I need to code to use MCP servers?
Not to use them. Installing from the desktop app's extensions directory needs no coding; writing your own server does.
Where do I run the claude mcp add command?
In your terminal, not inside a Claude session — you're configuring before starting a conversation.
What's the difference between local and remote servers?
Remote servers are reached over a URL. Local servers run as a subprocess on your machine, which lets them access local resources like your filesystem.

MCP works on any paid plan

Connecting Claude to your own tools is included — no add-on required.


Related