This article is compiled by the AIbase editing team

This is the fifth article in the MCP client getting-started tutorial series: "MCP Client Getting-Started Tutorial: Installation and Configuration of Claude".

This tutorial introduces how to install Trae IDE and configure MCP services to help you quickly master the combined use of Claude and MCP, and improve AI programming efficiency.

MCP Client Getting-Started Tutorial: Installation and Configuration of Claude

AIbase is glad to bring you this MCP client getting-started tutorial, which mainly introduces the installation and configuration process of the Claude desktop version. Through this tutorial, you will learn how to install the Claude Desktop client and configure the MCP server to expand the functions of the AI assistant.

What is MCP?

MCP (Model Context Protocol) is an open protocol for standardizing how applications provide context to large language models (LLMs). Through the MCP protocol, AI models can seamlessly integrate with external data sources and tools to access real-time data and perform more complex tasks.

The core components of MCP include:

• Host: The application (such as Claude Desktop) that runs the LLM, responsible for initiating the connection with the MCP server.
• Client: Runs inside the host application and establishes a 1:1 connection with the MCP server.
• Server: Provides access to external data sources and tools and responds to client requests.
• LLM: A large language model that obtains context through MCP and generates output.

• MCP enhances the functions of LLMs through three main primitives:

1. Prompts: Predefined instructions or templates that guide the LLM on how to process input or generate output.
2. Resources: Structured data that provides additional context, such as file or database contents.
3. Tools: Executable functions that allow the LLM to perform operations (such as querying an API) or retrieve information.

Detailed steps:

1. Download and install the Claude desktop client
2. Visit the official Claude download page: https://claude.ai/download
3. Select the download according to your operating system: For Windows users: Click "Download for Windows". For macOS users: Click "Download for Mac".

Note: Currently, Claude Desktop does not support the Linux system.

Install the Claude desktop client

Windows installation steps:

1. After the download is complete, find the downloaded installation file (usually in.exe format)
2. Double-click the installation file to start the installer
3. Follow the instructions of the installation wizard to complete the installation
4. After the installation is complete, Claude Desktop will start automatically

macOS installation steps:

1. After the download is complete, find the downloaded.dmg file
2. Double-click to open the.dmg file
3. Drag the Claude icon to the Applications folder
4. Launch Claude from the Applications folder or Launchpad

Update the Claude desktop version

AIbase warmly reminds you that if you have already installed Claude Desktop, you can check for updates through the following steps:

1. Open the Claude Desktop application
2. Click the Claude menu at the top
3. Select "Check for updates"
4. If there is an update, follow the prompts to complete the update process
5. Configure the MCP server - File system function
6. In this section, we will configure Claude Desktop to use the file system MCP server, which will enable Claude to read, write, and manage files on your computer.

Prerequisites

• The Claude Desktop client is installed
• Node.js is installed (for running the MCP server)
• If you haven't installed Node.js yet, please visit the official Node.js website to download and install it first.

Install Node.js

Before continuing to configure the MCP server, please make sure that Node.js is installed on your system:

1. Open the command line (cmd or PowerShell for Windows, Terminal for macOS)
2. Enter the following command to verify whether Node.js is installed:

node --version

1. If the version number is displayed, it means that it is correctly installed; otherwise, please go to the official Node.js website to download and install it.

Add file system MCP server configuration

1. Open Claude Desktop on your computer

2. Click the Claude menu and select "Settings..."

3. Click the "Developer" tab on the left side of the settings window, and then click the "Edit Config" button

4. Find the JSON file of Claude and double-click to enter the settings

5. Add the MCP server. For example, if you want to call the MCP of Conch AI, you just need to paste this configuration code here

Restart Claude

After configuring the MCP server, you need to restart Claude Desktop to apply the changes:

1. Completely close the Claude Desktop application
2. Reopen Claude Desktop

Verify the MCP server configuration

After restarting, you should see the icon of the file system tool in the Claude interface:

1. In the lower right corner of the Claude dialog box, a hammer icon should appear

2. Click the hammer icon, and you should see a list of tools provided by the file system MCP server

3. Then you can see the Conch added by AIbase here in the MCP

Simple example of the MCP Server -

The following is a simple example of a text-to-audio MCP server:

AIbase just enters your requirements in the input box, such as: Convert the provided content into audio: [Enter the text you want to convert to audio here], and then press Enter.

After a while, Claude will save the converted audio locally. The effect is as follows:

Summary

Through this tutorial, you have learned how to download and install the Claude Desktop client and configure the file system MCP server, enabling Claude to access and operate files on your computer.

The MCP protocol greatly expands the capabilities of the AI assistant, allowing it to interact with external systems and data. As you gain a deeper understanding of MCP, you can explore more servers and even create your own MCP server to further customize the functions of Claude.

We hope this tutorial is helpful to you! If you have any questions, please refer to the official MCP documentation or seek support in the community.

About the author: This tutorial is written by the AIbase editing team, which is committed to providing high-quality AI tool usage tutorials.