Managing servers usually means opening the dashboard, finding the right site, clicking through a few screens, and repeating that every time you need something. xCloud MCP removes that step.

MCP (Model Context Protocol) is an open standard that lets AI agents connect to external tools and act on your behalf. With xCloud MCP connected, you can simply ask your AI agent to check a server’s health, run a backup, update plugins, purge a cache, or pull the latest vulnerability scan, and it will do it through your xCloud account. No dashboard, no tab switching.

Follow the guide walks you through connecting xCloud MCP to AI Agents. 

Step 1: Open the MCP page in your xCloud dashboard #

From your xCloud dashboard, go to Settings. In the left sidebar, choose Developers → MCP.

This opens the MCP documentation, where you will find setup instructions for each supported AI agent.

Step 2: Copy the MCP URL #

We are using Claude Desktop for this walkthrough. Open the Claude Desktop section in the documentation and copy the MCP URL shown there.

Step 3: Add the connector in Claude Desktop #

Then Login to your Claude Desktop account and find profile and click Settings option from the menu.

In the popup, go to Connectors. Click Add, then choose Add Custom Connector from the dropdown.

Next, enter a name for the connector and paste the MCP URL you copied. Click Add. 

Claude Desktop will then show a screen asking you to connect xCloud MCP. Click on it.

Step 4: Authorize the connection #

The xCloud authorization page opens. Here you can:

• Verify the account email shown.
• Check which team the MCP is connecting to. If it is not the right one, click Switch to change it.
• Choose the access level: Full Access (read and write) or Read Only.

If you are only trying this out, Read Only is a safe place to start. You can reconnect with wider access later. Now, click Authorize to finish.

Step 5: Set tool permissions #

Once the connection is live, open the Permissions section. Every xCloud tool can be set individually to:

• Always Allow: the agent runs it without asking.
• Needs Approval: the agent asks you before each run.
• Blocked: the agent cannot use it at all.

You can also configure custom permissions per tool. A good default is Always Allow for read actions like listing sites or checking monitoring stats, and Needs Approval for anything that changes your infrastructure, such as deleting a site or restarting a service.

You are connected #

That is it. xCloud MCP is now connected to Claude Desktop, and your agent can work directly with your servers and sites. Try starting with something simple to confirm it is working:

• “List all my xCloud servers.”
• “Which of my sites have pending WordPress updates?”
• “Take a backup of my staging site.”

From here, you can connect the same MCP to other agents you use, adjust permissions as you get comfortable, and let your agent handle the routine infrastructure work while you focus on everything else.

Connecting xCloud MCP takes a couple of minutes, and it changes how you work with your infrastructure. Start small to build confidence, ask it to list your servers, check which sites have pending updates, or take a backup of a staging site, then widen the permissions as you get a feel for it.