MCPLab: Guide MCP Server Deployment to Integrate Microsoft Copilot Studio for Real - time Model

Mcplab

This experiment guides users on how to deploy the MCP server and integrate it into Microsoft Copilot Studio. Connect AI models with data sources through a standardized protocol to achieve real - time data access and tool integration.

Artificial intelligence chatbots Developer tools #AI integration #Data protocol #Real - time access #Tool connection .TypeScript

rating : 2 points

downloads : 7.4K

update time : 2025-07-25

Open Site

What is MCP?

The Model Context Protocol (MCP) is an open - source protocol defined by Anthropic that standardizes how applications provide context to LLMs. MCP provides a standardized way to connect AI models to different data sources and tools.

How to use the MCP server?

By deploying the MCP server and adding it as a tool in Copilot Studio, achieve real - time data access and enhance AI capabilities.

Applicable scenarios

Suitable for scenarios where existing knowledge servers and APIs need to be directly integrated into Copilot Studio, especially in situations that require enterprise - level security controls.

Main Features

Standardized integration

Provide a standardized way to connect AI models with data sources/tools

Enterprise security

Support enterprise security controls such as virtual network integration and data loss prevention

Real - time data access

Provide real - time data access capabilities for AI - driven agents

Advantages

Seamlessly collaborate with the connector infrastructure

Support multiple enterprise security controls

Provide real - time data access capabilities

Limitations

Currently only support the Tools function

After deployment, it is publicly accessible by default (manual closure is required)

How to Use

Prepare the environment

Install necessary tools such as VS Code, Node.js, and Git

Deploy the server

You can choose to run it locally or deploy it to Azure

Integrate into Copilot Studio

Import the MCP server through a custom connector

Create an agent and add tools

Create a new agent in Copilot Studio and add the MCP server as a tool

Usage Examples

Get a joke

Get a specific type of joke through the MCP server

Development scenario

Use GitHub Copilot in VS Code to access the MCP server

Frequently Asked Questions

Will MCP replace connectors?

How to ensure security after deploying to Azure?

Why do I see the 'Method not allowed' error?

Related Resources

Introduction to the MCP protocol

Official introduction to the Model Context Protocol

Azure Developer CLI

Installation guide for the Azure development command - line tool

Known issues and improvement plans for MCP

Known issues and future improvements of MCP in Microsoft Copilot Studio

🚀 Microsoft Copilot Studio ❤️ MCP

Welcome to the Microsoft Copilot Studio ❤️ MCP lab. In this lab, you'll learn how to deploy an MCP Server and add it to Microsoft Copilot Studio.

🚀 Quick Start

✨ Features

Standardized Protocol: Model Context Protocol (MCP) is an open protocol defined by Anthropic, which standardizes how applications provide context to LLMs. It offers a standardized way to connect AI models to different data sources and tools, enabling makers to integrate existing knowledge servers and APIs directly into Copilot Studio.
Better with Connectors: MCP servers use connector infrastructure, allowing them to employ enterprise security and governance controls while supporting real - time data access for AI - powered agents.

📦 Installation

⚙️ Prerequisites

Visual Studio Code (link)
Node v22 (ideally installed via nvm for Windows or nvm)
Git installed (link)
Azure Developer CLI (link)
Azure Subscription (with payment method added)
GitHub account
Copilot Studio trial or developer account
Power Platform environment provisioned - with the following toggle on:

➕ Create a new GitHub repository based on the template

Select Use this template.
Select Create a new repository.
Select the right Owner (it might already be selected when you have only one owner to choose from).
Give it a Repository name.
Optionally, give it a Description.
Select Private.
Select Create repository.

This will take a little while. After it's done, you will be directed to the newly created repository.

⚖️ Choice: Run the server locally or deploy to Azure

Now you have a choice! You can either run the server locally or deploy it to Azure. There are a couple of steps that you need to do for both:

Clone this repository by running the following command (replace {account} by your own GitHub account name):

git clone https://github.com/{account}/mcsmcp.git
Open Visual Studio Code and open the cloned folder.
Open the terminal and navigate to the cloned folder.

🏃‍♀️ Run the MCP Server Locally

Run npm install.
Run npm run build && npm run start.
Select PORTS at the top of the Visual Studio Code Terminal.
Select the green Forward a Port button.
Enter 3000 as the port number (this should be the same as the port number you see when you ran the command in step 5). You might be prompted to sign in to GitHub, if so please do this, since this is required to use the port forwarding feature.
Right click on the row you just added and select Port visibility > Public to make the server publicly available.
Ctrl + click on the Forwarded address, which should be something like: https://something-3000.something.devtunnels.ms.
Select Copy on the following pop - up to copy the URL.
Open to the browser of your choice and paste the URL in the address bar, type /mcp behind it and hit enter.

If all went well, you will see the following error message:

{"jsonrpc":"2.0","error":{"code":-32000,"message":"Method not allowed."},"id":null}

🌎 Deploy to Azure

⚠️ Important Note

As listed in the prerequisites, the Azure Developer CLI needs to be installed on your machine for this part.

Make sure to login to Azure Developer CLI if you haven't done that yet.

azd auth login

⚠️ Important Note

After running azd up, you will have an MCP Server running on Azure that is publicly available. Ideally, you don't want that. Make sure to run azd down after finishing the lab to delete all the resources from your Azure subscription. Learn how to run azd down by going to this section.

Run the following command in the terminal:

azd up

For the unique environment name, enter mcsmcplab or something similar. Select the Azure Subscription to use and select a value for the location. After that, it will take a couple of minutes before the server has been deployed. When it's done - you should be able to go to the URL that's listed at the end and add /mcp to the end of that URL.

You should again see the following error:

{"jsonrpc":"2.0","error":{"code":-32000,"message":"Method not allowed."},"id":null}

💻 Usage Examples

👨‍💻 Use the Jokes MCP Server in Visual Studio Code / GitHub Copilot

To use the Jokes MCP Server, you need to use the URL of your server (can be either your devtunnel URL or your deployed Azure Container App) with the /mcp part at the end and add it as an MCP Server in Visual Studio Code.

Press either ctrl + shift + P (Windows/Linux) or cmd + shift + P (Mac) and type MCP.
Select MCP: Add Server....
Select HTTP (HTTP or Server - Sent Events).
Paste the URL of your server in the input box (make sure /mcp in the end is included).
Press Enter.
Enter a name for the server, for instance JokesMCP.
Select User Settings to save the MCP Server settings in your user settings.

This will add an MCP Server to your settings.json file. It should look like this:
Open GitHub Copilot.
Switch from Ask to Agent.
Make sure the JokesMCP server actions are selected when you select the tools icon:

Ask the following question:

Get a chuck norris joke from the Dev category

This should give you a response like this:

Now you have added the JokesMCP server to Visual Studio Code!

👨‍💻 Use the Jokes MCP Server in Microsoft Copilot Studio

Import the Connector

Go to https://make.preview.powerapps.com/customconnectors (make sure you’re in the correct environment) and click + New custom connector.
Select Import from GitHub.
Select Custom as Connector Type.
Select dev as the Branch.
Select MCP - Streamable - HTTP as the Connector.
Select Continue.
Change the Connector Name to something appropriate, like for instance Jokes MCP.
Change the Description to something appropriate.
Paste your root URL (for instance something-3000.something.devtunnels.ms or something.azurecontainerapps.io) in the Host field.
Select Create connector.

⚠️ Important Note

You may see a warning and an error upon creation – it should be resolved soon - but you can ignore it for now.

Close the connector.

Create an agent and add the MCP server as a tool

Go to https://copilotstudio.preview.microsoft.com/.
Select the environment picker at the top right corner.
Select the right environment (the environment with the Get new features early toggle switched on).
Select Create in the left navigation.
Select the blue New agent button.
Select the Configure tab on the left.
Change the name to Jokester.

Add the following Description

A humor - focused agent that delivers concise, engaging jokes only upon user request, adapting its style to match the user's tone and preferences. It remains in character, avoids repetition, and filters out offensive content to ensure a consistently appropriate and witty experience.

Add the following Instructions

You are a joke - telling assistant. Your sole purpose is to deliver appropriate, clever, and engaging jokes upon request. Follow these rules:

* Respond only when the user asks for a joke or something related (e.g., "Tell me something funny").
* Match the tone and humor preference of the user based on their input—clean, dark, dry, pun - based, dad jokes, etc.
* Never break character or provide information unrelated to humor.
* Keep jokes concise and clearly formatted.
* Avoid offensive, discriminatory, or NSFW content.
* When unsure about humor preference, default to a clever and universally appropriate joke.
* Do not repeat jokes within the same session.
* Avoid explaining the joke unless explicitly asked.
* Be responsive, witty, and quick.

Select Continue on the top right.
Enable Generative AI Orchestration.
Disable general knowledge in the Knowledge section.
Select Tools in the top menu.
Select Add a tool.
Select the Model Context Protocol tab to filter all the Model Context Protocol Servers (see number 1 in the screenshot below).
Select the Jokes MCP server (see number 2 in the screenshot below).
Create a new connection by selecting the Not connected and Create new Connection.
Select Create.
Select Add to agent to add the tool to the agent.
Select the refresh icon in the Test your agent pane.
In the Test your agent pane send the following message:
```
Can I get a Chuck Norris joke?
```
This will show you message that additional permissions are required to run this action. This is because of the user authentication in the action wizard.
Select Connect.

This will open a new window where you can manage your connections for this agent.
Select Connect next to the JokesMCP.
Wait until the connection is created and select Submit.
The connection should now be connected, so the status should be set to Connected.
Close the manage your connections tab in your browser.

Now you should be back in the Jokester agent screen.
Select the refresh icon in the Test your agent pane.
In the Test your agent pane send the following message:
```
Can I get a Chuck Norris joke?
```
This will now show a Chuck Norris joke - instead of the additional permissions. If that's not the case - you probably have missed the prerequisite that the environment needs to have the get new features early toggle on.
In the Test your agent pane send the following message:
```
Can I get a Dad joke?
```
This will now show a Dad joke.

And that was the Jokes MCP Server working in Microsoft Copilot Studio.

❌ Remove the Azure resources

To remove the Azure resources after finishing the lab, run the following command in the terminal:

azd down

This command will show you the resources that will be deleted and then ask you to confirm. Confirm with y and the resources will be deleted. This can take a couple of minutes, but at the end you will see a confirmation:

💡 Known issues and planned improvements

There are some known issues and planned improvements for MCP in Microsoft Copilot Studio. They are listed in this Microsoft Learn article.

🗣️ Feedback

Hopefully you liked the lab. Please take the time to fill in our feedback form to tell us how we can improve!

🚀 Contributing

This project welcomes contributions and suggestions. Most contributions require you to agree to a Contributor License Agreement (CLA) declaring that you have the right to, and actually do, grant us the rights to use your contribution. For details, visit https://cla.opensource.microsoft.com.

When you submit a pull request, a CLA bot will automatically determine whether you need to provide a CLA and decorate the PR appropriately (e.g., status check, comment). Simply follow the instructions provided by the bot. You will only need to do this once across all repos using our CLA.

This project has adopted the Microsoft Open Source Code of Conduct. For more information see the Code of Conduct FAQ or contact opencode@microsoft.com with any additional questions or comments.

™️ Trademarks

This project may contain trademarks or logos for projects, products, or services. Authorized use of Microsoft trademarks or logos is subject to and must follow Microsoft's Trademark & Brand Guidelines. Use of Microsoft trademarks or logos in modified versions of this project must not cause confusion or imply Microsoft sponsorship. Any use of third - party trademarks or logos are subject to those third - party's policies.