Telerik Reporting MCP Server

The Telerik Reporting MCP (Model Context Protocol) server enhances your AI-powered development experience by providing specialized context about Telerik Reporting.

This MCP server enables AI-powered IDEs and tools to generate more accurate, tailored code that leverages Telerik Reporting APIs.

You can ask complex questions about integrating Telerik Reporting in web applications, request specific implementations, and generate comprehensive code solutions.

Prerequisites

To use the Telerik Reporting MCP Server, you need:

• .NET 8+ SDK.
• A Telerik user account.
• An active license covering the usage of Telerik Reporting.
• An application that uses the Telerik Reporting.
• An MCP-compatible client (IDE, Code Editor, or App) that supports MCP tools. For example, the latest Visual Studio 2026.

Installation

Depending on your environment, you can install the Telerik Reporting MCP server in any of the following ways:

bash

dnx Telerik.Reporting.MCP

bash

dotnet tool install -g Telerik.Reporting.MCP

Next, make sure the configuration in your mcp.json is correct, and then add your Telerik license.

Configuring mcp.json

Use the settings in the following table to configure the Telerik Reporting MCP server in the mcp.json file of your code editor. Select the correct value based on your development environment.

Setting Name	.NET 10 Value	.NET 8/9 Value
Package Name	"Telerik.Reporting.MCP"	"Telerik.Reporting.MCP"
Type	"stdio"	"stdio"
Command	"dnx"	"dotnet"
Arguments	["Telerik.Reporting.MCP", "--yes"]	["tool", "run", "telerik-reporting-assistant"]
Server Name	"telerik-reporting-assistant"	"telerik-reporting-assistant"

Configuring Your License

An active Telerik Reporting license is required to use the Telerik Reporting MCP server.

• When installing the MCP server by using the .NET tooling (dnx or dotnet tool install), the license key file will be retrieved automatically if it is present in the default directory on your system (%AppData%\Telerik\telerik-license.txt on Windows and ~/.telerik/telerik-license.txt on Linux).

• When using the .NET tooling, if your license key file is not in the default directory, you can add the license key file as an environment variable in the mcp.json file using one of the following options:

  JSON

  "env": {
  	"TELERIK_LICENSE_PATH": "THE_PATH_TO_YOUR_LICENSE_FILE"
  }

  JSON

  "env": {
  	"TELERIK_LICENSE": "YOUR_LICENSE_KEY_HERE"
  }

Using a license file path is recommended unless you are sharing settings across different systems. Remember to update the license key when renewing.

Visual Studio Configuration

• Early Visual Studio 17.14 versions require the Copilot Chat window to be open when a solution loads for the MCP server to work properly.
• For complete setup instructions, see Use MCP servers in Visual Studio.

The Telerik Reporting MCP server is available as a NuGet package: https://www.nuget.org/packages/Telerik.Reporting.MCP.

Automatic Configuration through the Visual Studio UI

1. Open the Configure MCP server menu in Visual Studio

   • Open the Visual Studio GitHub Copilot Chat
   • Click on the button at the bottom right corner to open the Select tools context menu.
   • Click on the + button at the top right corner of the menu to open the Configure MCP server menu.

   

2. Use the following values for the configuration fields:

   • In Destination select either Global-Available in all solutions or Solution-Available in this solution.
   • For Server ID use telerik-reporting-assistant.
   • Select stdio as Type.
   • The Command (with optional arguments) field should contain dnx Telerik.Reporting.MCP --yes.
   • Click on the Save button to preserve the configuration.

   

3. Enable the telerik-reporting-assistant tool in the Copilot Chat window's tool selection dropdown.

   The telerik-reporting-assistant tool should now appear in the Select tools context menu. Check it to enable the Telerik Reporting AI Coding Assistant:

   

Manual Configuration

The steps below describe the sample procedure for configuring the Telerik Reporting MCP server in Visual Studio.

1. Add an .mcp.json file to either of the following locations:

   • For a workspace-specific setup, add the file to the solution's folder.
   • For a global setup, add the file to your user directory, %USERPROFILE% (for example, C:\Users\YourName\.mcp.json).

2. Add the following configuration to the .mcp.json file:

   JSON

   {
   	"servers": {
   		"telerik-reporting-assistant": {
       		"type": "stdio",
       		"command": "dnx",
       		"args": ["Telerik.Reporting.MCP", "--yes"]
   		}
   	}
   }

   JSON

   {
   	"servers": {
   		"telerik-reporting-assistant": {
       		"type": "stdio",
       		"command": "dotnet",
       		"args": ["tool", "run", "telerik-reporting-assistant"]
   		}
   	}
   }

3. Restart Visual Studio.

4. Enable the telerik-reporting-assistant tool in the Copilot Chat window's tool selection dropdown.

Known Issue: Hanging tool calls in Visual Studio

When using Telerik AI tools in Visual Studio, GitHub Copilot may:

• hang during tool invocation;
• show UI for a successful tool response, but actually fail silently;
• continue generation without waiting for parallel tool calls.

In these cases, the response may be generated but not provided to the Copilot Agent UI.

This is a known issue in Visual Studio Copilot, not related to Telerik MCP servers or AI tools, and does not reproduce in VS Code.

For more details, see the related Visual Studio Developer Community issue:
https://developercommunity.visualstudio.com/t/Copilot-stopped-working-after-latest-upd/10936456

Microsoft has acknowledged the issue and marked it as Fixed - Pending Release. A future Visual Studio update is expected to resolve it.

Telerik Visual Studio Extension Setup:

The .mcp.json file can be created and updated by using the Telerik Visual Studio Extension menu in Visual Studio.

The Configure MCP Server Globally option creates (or updates if existing) the global %USERPROFILE%\.mcp.json file and adds the Telerik Reporting server entry.

The Configure MCP Server for Solution option is visible only if you have a project opened in Visual Studio. The option creates (or updates if existing) the .mcp.json file in the solution directory.

The Add/Update GitHub Copilot Instructions option is visible only if you have a project opened in Visual Studio. The option creates (or updates if existing) a copilot-instructions.md file in the .github\ folder of the solution (\SolutionDir\.github\copilot-instructions.md), which includes extra context that is automatically used with the prompt given to the Copilot chat in Visual Studio.

Visual Studio Usage

After configuration and restart:

1. Open Copilot Chat.

2. Enable the telerik-reporting-assistant tool.

3. Grant permissions when prompted (per session, workspace, or always).

   

Start fresh sessions for unrelated prompts to avoid context pollution. You can check the Output pane of Visual Studio for diagnostics (select output from GitHub Copilot).

Visual Studio Code

• Visual Studio Code 1.102.1 or later is required to use the Telerik Reporting MCP Server.
• For complete setup instructions, see Use MCP servers in Visual Studio Code.

The basic setup in Visual Studio Code involves the following steps:

1. Enable chat.mcp.enabled in Visual Studio Code settings.
2. Create .vscode/mcp.json in your workspace root (or user folder for global setup).

The settings below allow you to use the #telerik-reporting-assistant handle to call the Telerik Reporting MCP server. To use a custom handle, change the server name in the mcp.json.

JSON

{
	"servers": {
		"telerik-reporting-assistant": {
			"type": "stdio",
			"command": "dnx",
			"args": ["Telerik.Reporting.MCP", "--yes"]
		}
	}
}

JSON

{
	"servers": {
		"telerik-reporting-assistant": {
			"type": "stdio",
			"command": "dotnet",
			"args": ["tool", "run", "telerik-reporting-assistant"]
		}
	}
}

1. For global discovery, enable chat.mcp.discovery.enabled in settings.json:

   JSON

   {
     "chat.mcp.discovery.enabled": true
   }

2. Restart Visual Studio Code.

   

Cursor

For complete setup instructions, see Model Context Protocol.

Create a .cursor/mcp.json file in your workspace root (or user folder for global setup):

JSON

{
"mcpServers": {
	"telerik-reporting-assistant": {
			"type": "stdio",
			"command": "dnx",
			"args": ["Telerik.Reporting.MCP", "--yes"]
		}
	}
}

JSON

{
	"mcpServers": {
		"telerik-reporting-assistant": {
			"type": "stdio",
			"command": "dotnet",
			"args": ["tool", "run", "telerik-reporting-assistant"]
		}
	}
}

This enables you to call the MCP Server with the #telerik-reporting-assistant handle. To use a custom handle, change the server name in the mcp.json.

Usage

When switching between tasks and files, start a new session in a new chat window to avoid polluting the context with irrelevant or outdated information.

To use the Telerik Reporting MCP Server:

1. Choose your preferred mode and model.
   At the time of publishing, Claude Sonnet 4 and GPT-5 produce optimal results.

2. Start your prompt with #telerik-reporting-assistant (or with '#' followed by your custom MCP server name, if set):

3. Inspect the output and verify that the MCP server is used. Look for messages similar to the ones below:

   • Visual Studio: Running telerik-reporting-assistant
   • Visual Studio Code: Running telerik-reporting-assistant
   • Cursor: Calling MCP tool telerik-reporting-assistant

4. If the Telerik Reporting MCP server is not used even though it's installed and enabled, double-check the server name in your configuration and try rephrasing your prompt.

5. If requested, grant the Telerik Reporting MCP Server permissions to run for this session, workspace, or always.

6. Start fresh sessions for unrelated prompts to avoid context pollution.

Improving Server Usage

Determining the Reporting Area: To better specify the context for each prompt, the MCP Server will attempt to determine the Telerik Reporting Area the prompt relates to. These areas represent the supported web frameworks and backend technologies, as listed below as pairs consisting of an identifier and a description:

• AngularWrapper - Angular Report Viewer (wrapper of the HTML5 Report Viewer)
• NativeAngular - Native Angular Report Viewer (built with Kendo UI for Angular)
• BlazorWrapper- Blazor Report Viewer (wrapper of the HTML5 Report Viewer)
• NativeBlazor - Native Blazor Report Viewer (built with Telerik UI for Blazor)
• HTML5 - HTML5 Report Viewer (built with Kendo UI for jQuery)
• React - React Report Viewer (wrapper of the HTML5 Report Viewer)
• RESTService - Reporting REST Service for .NET Framework
• RESTServiceCore - Reporting REST Service for .NET Core/.NET 8+
• General - General question

If the Reporting Area cannot be determined automatically, the MCP Server will display a dialog asking to pick the corresponding area in which the prompt falls in:

After clicking on the 'Respond' button, the MCP Server will display a drop-down list with all the areas and wait for selecting the corresponding Reporting Area. If there is no applicable entry, please select the 'General' entry at the end of the list.

Custom Instructions: To increase the likelihood of the Telerik MCP server being used, or to call it without the need to mention "telerik-reporting" explicitly, add custom instructions to your AI tool:

• GitHub Copilot custom instructions
• Cursor rules

Sample Prompts

The following examples demonstrate useful prompts for the Telerik Reporting MCP Server:

• "#telerik-reporting-assistant Add a Native Angular Report Viewer to my Angular application."
• "#telerik-reporting-assistant Connect the Telerik Report Viewer to my Telerik Report Server instance."

Visit the Telerik Reporting Prompt Library for more sample prompts.

Usage Limits

A Telerik Subscription or Trial license is required to use the Telerik Reporting AI Coding Assistant.

Perpetual licenses do not include access, but you can start a free trial to evaluate it.

See the Usage Limits section for details.

Connect to Local AI Model

You can use the Telerik Reporting MCP server with local large language models (LLMs):

1. Run a local model, for example, through Ollama.
2. Use a bridge package like MCP-LLM Bridge.
3. Connect your local model to the Telerik MCP server.

This setup allows you to use the Telerik AI Coding Assistant without cloud-based AI models.

See Also

• AI Coding Assistant Overview
• Telerik Reporting Prompt Library
• Telerik Blazor MCP Server