MCP Server for Red Hat OpenShift AI

This project implements a Model Context Protocol (MCP) server for managing Red Hat OpenShift AI (RHOAI) workbenches. It provides tools for listing, creating, and managing workbenches, images, hardware profiles, and storage in OpenShift clusters, also it provides tools for monitoring resource consumption by workbenches per user, workbench, namespace or cluster. This project is part of a Bachelor's thesis.

Prerequisites

• Go 1.24.0 or later
• Access to an OpenShift cluster with RHOAI installed
• oc CLI logged in to your cluster
• (Optional) golangci-lint for code linting

Installation

1. Clone the repository:

git clone https://github.com/amaly/mcp-server-rhoai.git
cd mcp-server-rhoai

2. Download dependencies:

go mod download

3. Ensure you're logged in to your OpenShift cluster:

oc login <your-cluster-url>

Building

The project includes a Makefile for easy building:

Build with Make

# Build with linting and testing (recommended)
make build

# Build without running lint and tests (faster)
go build -o mcp-server-rhoai

The build process will:

1. Run linting checks (golangci-lint)
2. Execute all tests
3. Compile the binary as mcp-server-rhoai

Build Outputs

• Binary location: ./mcp-server-rhoai

Testing

The project has a comprehensive unit test suite (60 tests across 7 test files) covering every tool module. Tests use Kubernetes fake clients (dynamicfake.NewSimpleDynamicClient) and function-variable mocking to isolate logic from live cluster dependencies, so no running cluster is needed.

Run tests using Make or directly with Go:

# Run all tests with Make
make test

# Run tests directly with verbose output
go test -v ./tools/... ./resources/... ./prompts/...

Test coverage includes:

• Workbench tools — CRUD operations, status transitions, and helper functions (image tag parsing, resource extraction, PVC name resolution)
• Storage tools — PVC creation, deletion, and size updates including rejection of size decreases
• Image tools — Custom image CRUD, default-image and in-use protection checks, image listing
• Hardware profile tools — Profile CRUD and resource identifier parsing
• Resource consumption tools — Value parsing (millicores, Gi suffixes, decimals), aggregation, and per-workbench/namespace/user/cluster queries
• Namespace & pod tools — Listing and filtering by namespace

Each tool module has a dedicated *_test.go file with both success and error-path cases. Table-driven subtests are used where multiple input variations need to be verified.

Evaluation

Uses Promptfoo for testing tool selection accuracy.

Run Evaluation

export GOOGLE_API_KEY=your-key-here
make eval        # run tests
make eval-view   # run and open results in browser

What It Tests

• Tool selection - correct tool called for each prompt
• Parameter extraction - arguments passed correctly (namespace, workbench name, etc.)

Key Learnings

Tool descriptions must be explicit for model-agnostic compatibility:

• Mark optional fields with omitempty and describe defaults in jsonschema descriptions
• Include valid values in descriptions (e.g., "status: 0 for running, 1 for stopped")
• Simpler models (Gemini Flash) are stricter than Claude - good for testing robustness

Configuration

The MCP server can be configured in AI assistants to enable workbench management through natural language.

VSCode/Cursor Setup

1. Build the server:

make build

2. Open MCP Settings in Cursor:

   • Press Ctrl + Shift + P (Windows/Linux) or Cmd + Shift + P (Mac)
   • Type "Open MCP Settings"
   • Click "New MCP server"

3. Add configuration:

Read-only mode (default, safer):

{
  "mcpServers": {
    "mcp-server-rhoai": {
      "command": "/home/amaly/mcp-server-rhoai/mcp-server-rhoai"
    }
  }
}

**Read-write

... View full README on GitHub