MCP Server for RHOAI Collaboration with Cursor IDE Guide

MCP Server Rhoai

This project is a bachelor's thesis project for implementing an MCP server for collaborating with RHOAI, providing a guide for setting up and using the MCP server in the Cursor IDE.

Developer tools Cloud platform #MCP Server #RHOAI Integration #Cursor Plugin #Go Language Development .Go

rating : 2 points

downloads : 5.9K

update time : 2025-12-03

Open Site

What is the RHOAI MCP Server Suite?

This is a set of tools developed based on the Go language, which follow the Model Context Protocol (MCP) standard. Simply put, they are like 'interpreters' or 'assistants', allowing you to directly use natural language (such as 'Show me all the workbenches') to query and manage resources on your Red Hat OpenShift AI cluster in an AI integrated development environment like Cursor, without leaving the editor to perform complex command - line operations.

How to use the RHOAI MCP Server Suite?

The usage process is divided into two steps: First, you need to compile the provided Go code into an executable program and register it as an MCP server in Cursor's settings. After the configuration is completed, you can directly ask questions to the AI assistant in Cursor's chat window. For example, 'List all the workbenches in the mcp - test namespace', and the AI assistant will automatically call the corresponding MCP tool to obtain information and answer you.

Applicable scenarios

When you are a developer working on an AI/ML project on the OpenShift AI platform, you need to frequently check the workbench status, Pod running status, or debug the environment. By using these MCP servers, you can quickly obtain cluster information while coding, without switching terminals or web consoles, which greatly improves work efficiency and context coherence.

Main Features

Workbench List Query

Provide a dedicated MCP server (such as WorkbenchList) for querying AI workbench resources in a specified OpenShift namespace, and return key information such as names and statuses.

Pod List Query

Provide a dedicated MCP server (such as podslist) for querying Pod (container group) resources in a specified OpenShift namespace, helping you understand the application running status.

Deep Integration with Cursor IDE

Follow the MCP protocol and can be seamlessly integrated into the Cursor IDE. After configuration, the AI assistant can directly call these tools to answer your questions about the cluster status.

Developed Based on Go Language

The server is written in Go language, compiled into an independent binary file, which is easy to deploy, runs efficiently, and has good cross - platform compatibility.

Cluster Authentication Required

To securely access the OpenShift cluster, you need to log in to the target cluster through methods such as `oc login` before using the relevant tools to ensure operation permissions.

Advantages

Improve development efficiency: You can query cluster information without leaving the IDE and switching contexts, allowing you to focus more on the code.

Natural language interaction: Manage the infrastructure through simple conversations with the AI assistant, reducing the threshold of using complex command lines.

Modular design: Each function (such as querying workbenches and querying Pods) is an independent server, which can be enabled and configured as needed.

Open - source and extensible: The code is open - source. You can easily develop MCP servers for querying other Kubernetes resources based on this project.

Limitations

Requires initial configuration: Users need to manually compile the code and configure the MCP server in Cursor, which has a certain technical threshold.

Depends on cluster login: The tool itself does not handle authentication. Users need to log in to the OpenShift cluster through the command line in advance.

Relatively basic functions: The current version mainly provides query functions, and complex cluster management operations (such as creation and deletion) may not be supported.

Specific to Cursor: Although MCP is a standard protocol, the configuration instructions in this README are mainly for the Cursor IDE.

How to Use

Prepare the Environment and Code

Ensure that the Go language environment is installed on your development machine. Clone or download the source code of this project from the code repository.

Compile the Server

Open the terminal, enter the project directory, and use the Go compilation command to build the source code into an executable binary file. You can name the output file whatever you like.

Configure Cursor MCP Settings

Open the Cursor IDE, press `Ctrl + Shift + P` (Windows/Linux) or `Cmd + Shift + P` (Mac), enter and select 'Open MCP Settings'. In the opened JSON configuration file, add the path of the compiled server.

Open the terminal and use the OpenShift command - line tool `oc` to log in to the target cluster you want to query. This is a prerequisite for the server to access cluster data.

Start Using

After the configuration is completed, in the AI chat interface of Cursor, you can directly ask questions to the assistant. The assistant will recognize your intention and automatically call the configured MCP tool to get the answer.

Usage Examples

Quickly Check Workbench Status

You are developing a machine - learning model and need to confirm whether the workbench providing you with the computing environment is ready, rather than stuck in the startup process.

Debug Application Deployment Issues

You just deployed a new service to OpenShift, but you can't access it. You want to quickly check whether the relevant Pods are running and if there are any errors.

Daily Environment Inspection

Before starting work every day, you want to quickly understand the basic situation of resources in several development environments you are responsible for.