openshift
OpenShift MCP Server - Monitor and manage OpenShift/Kubernetes clusters through AI IDE
claude mcp add --transport stdio redhat-performance-openshift-mcp-server node server.js \ --env MCP_SSH_KEY="SSH private key file path for key-based authentication (e.g., ~/.ssh/id_rsa_bastion)" \ --env MCP_REMOTE_PATH="remote path to MCP server on bastion (e.g., /opt/openshift-mcp-server)" \ --env MCP_BASTION_HOST="bastion host address (optional; used for remote launches)" \ --env MCP_BASTION_USER="bastion username (optional; used for remote launches)" \ --env MCP_REMOTE_NODE_ENV="node environment (default: production)" \ --env MCP_BASTION_PASSWORD="bastion password (optional; used for password-based SSH)" \ --env MCP_REMOTE_KUBECONFIG="path to remote kubeconfig on bastion or cluster"
How to use
OpenShift MCP Server provides real-time cluster monitoring and management capabilities for OpenShift/Kubernetes environments. It exposes core health checks, performance metrics, resource issue detection, and deployment monitoring, along with remote execution and benchmarking tools designed for both single-cluster and multi-cluster scenarios. The server also supports advanced features like bastion-host remote execution, SSH integration, and automated benchmarking workflows, enabling AI assistants to query cluster health, performance trends, and configuration states in natural language. To interact with the server, start it through the configured node entry point (server.js) and connect via MCP clients or tooling that consume MCP-formatted data. The included remote-launcher tooling helps orchestrate remote tests and benchmarks across clusters through a bastion host when needed.
How to install
Prerequisites:
- Node.js 18+ installed on your system (and npm/yarn as needed)
- Access to an OpenShift/Kubernetes cluster
- Optional: bastion host access for remote deployments
Install steps:
-
Clone the MCP server repository git clone https://github.com/your-org/redhat-performance-openshift-mcp-server.git cd redhat-performance-openshift-mcp-server
-
Install dependencies npm install
or: yarn install
-
Configure environment variables (as needed)
- For local runs: create a .env file or export variables in your shell, e.g.:
export MCP_REMOTE_PATH="/opt/openshift-mcp-server"
export MCP_REMOTE_KUBECONFIG="/path/to/kubeconfig"
export MCP_REMOTE_NODE_ENV="production"
If using bastion remote execution:
export MCP_BASTION_HOST="your-bastion-host.com" export MCP_BASTION_USER="your-username" export MCP_SSH_KEY="~/.ssh/id_rsa_bastion"
- For local runs: create a .env file or export variables in your shell, e.g.:
export MCP_REMOTE_PATH="/opt/openshift-mcp-server"
export MCP_REMOTE_KUBECONFIG="/path/to/kubeconfig"
export MCP_REMOTE_NODE_ENV="production"
-
Start the MCP server node server.js
If a custom start script is used, adjust the command accordingly
-
(Optional) Run remote launcher to test remote deployment/benchmarks node ./scripts/remote-launcher.js
-
Verify connectivity and health endpoints as per your environment’s tooling.
Additional notes
Tips and common issues:
- Ensure Node.js 18+ is installed and your environment variables are correctly set for remote/bastion usage when applicable.
- If connecting to remote clusters, verify SSH access or key-based authentication is functioning before launching remote tests.
- For multi-cluster testing, use the Multi-Cluster Testing features to manage tests across environments, and ensure proper kubeconfig paths are available on the remote side.
- When using the remote launcher, you can specify environment details (host, user, kubeconfig) via command-line arguments or environment variables; CLI overrides take precedence over env vars.
- Keep MCP_REMOTE_NODE_ENV aligned with your deployment profile (production, staging, etc.) to ensure deterministic behavior of node processes and benchmarking tasks.
- Review logs for sensitive data redaction warnings to ensure credentials are not exposed in logs.
Related MCP Servers
zen
Selfhosted notes app. Single golang binary, notes stored as markdown within SQLite, full-text search, very low resource usage
MCP -Deepseek_R1
A Model Context Protocol (MCP) server implementation connecting Claude Desktop with DeepSeek's language models (R1/V3)
mcp-fhir
A Model Context Protocol implementation for FHIR
mcp
Inkdrop Model Context Protocol Server
mcp-appium-gestures
This is a Model Context Protocol (MCP) server providing resources and tools for Appium mobile gestures using Actions API..
dubco -npm
The (Unofficial) dubco-mcp-server enables AI assistants to manage Dub.co short links via the Model Context Protocol. It provides three MCP tools: create_link for generating new short URLs, update_link for modifying existing links, and delete_link for removing short links.