🎯 Kubernetes MCP Server
AI-powered MCP server understands natural language queries about your Kubernetes cluster
What is Kubernetes MCP Server?
Kubernetes MCP Server is an AI-powered control plane that allows users to manage their Kubernetes clusters through natural language queries.
How to use Kubernetes MCP Server?
To use the server, ensure your Kubernetes environment is configured, then interact with it by asking questions in natural language about your cluster's status and management.
Key features of Kubernetes MCP Server?
- AI-powered diagnostics and recommendations for pod management.
- Intelligent Helm chart management and release upgrades.
- Real-time event monitoring and bottleneck detection.
- Secure command execution within pods.
Use cases of Kubernetes MCP Server?
- Monitoring cluster health and diagnosing issues.
- Managing application deployments and troubleshooting pods.
- Performing Helm release management and tracking configurations.
FAQ from Kubernetes MCP Server?
- Can I use natural language to interact with the server?
Yes! The server is designed to understand natural language queries.
- What are the prerequisites for using the server?
You need JDK 17 or later, Gradle 7.x or later, and a configured Kubernetes environment.
- Is the server secure?
Yes! It includes features like secure command execution and namespace isolation.
🎯 Kubernetes MCP Server
Your AI-Powered Kubernetes Control Plane
⎈ K8s MCP Server
├── 🤖 AI-Powered
├── 🔍 Smart Diagnostics
├── 🛡️ Enhanced Security
└── 🚀 Developer Friendly
✨ Overview
This is a Spring Boot-based MCP server that combines the power of AI with cluster management capabilities.
🎁 Features
🔄 Pod Management
- 📋 List and analyze pods in real-time
- 📝 Smart log analysis with error pattern detection
- 🔍 AI-powered pod diagnostics with recommendations
- ⚡ Secure command execution in pods
🎯 Job Management
- 📋 List and analyze jobs in any namespace
- 🔍 Get detailed job status and execution history
- 🗑️ Clean up completed or failed jobs
⎈ Helm Integration
- 📦 Intelligent chart management
- 🔄 Seamless release upgrades
- 🗃️ Repository management
- 📊 Configuration tracking
📈 Event Analysis
- 🎯 Real-time event monitoring
- 🚨 Smart bottleneck detection
- 📱 Live deployment tracking
🛠️ Prerequisites
| Requirement | Version |
|---|---|
| ☕ JDK | 17 or later |
| 🐘 Gradle | 7.x or later |
| ⎈ Kubernetes | Configured ~/.kube/config |
| 🎡 Helm | CLI installed |
Note: MCP tool always uses the kubeconfig file from
~/.kube/config, so make sure it is properly configured.
🏗️ Building the Project
./gradlew clean build
🤝 Integration with Claude Desktop
- Install Claude Desktop
- Configure the MCP server connection:
{
"mcpServers": {
"spring-ai-mcp-k8s": {
"command": "java",
"args": [
"-Dspring.ai.mcp.server.stdio=true",
"-Dspring.main.web-application-type=none",
"-Dlogging.pattern.console=",
"-jar",
"<<jar-location>>"
]
}
}
}
Note: Make sure to pick the fat jar with all the dependencies
Integration with Other MCP Hosts
The server follows the standard MCP protocol and can be integrated with any MCP host that supports Spring-based MCP servers. Configure your host to point to the server's URL.
🗣️ Natural Language Interactions
💡 Just ask questions naturally - no need to memorize commands!
This AI-powered MCP server understands natural language queries about your Kubernetes cluster. Here are examples of questions you can ask:
🏥 Cluster Health and Diagnostics
📊 What's the overall health of my cluster?
🔍 Are there any resource bottlenecks in the 'production' namespace?
🚨 Show me problematic pods in the 'dev' namespace with recommendations
📅 What events happened in the cluster in the last hour?
📱 Application Management
📋 List all pods in the 'default' namespace and their status
❓ Why is the 'auth-service' pod failing to start?
📝 Show me the logs from the 'payment-processor' pod with error highlighting
📈 What's using the most resources in the 'monitoring' namespace?
🔍 Show me the status of all jobs in the 'default' namespace
🎯 List completed jobs in the 'data-processing' namespace
🗑️ Clean up failed jobs in the 'etl' namespace
⎈ Helm Release Management
📦 What Helm releases are installed in the 'staging' namespace?
⚡ Install the 'prometheus' chart from the official repository
⚙️ What values are configured for the 'elasticsearch' release?
🔄 Update the 'kafka' release to version 2.0.0
Note: I have noticed sometimes LLM will try and generate kubectl commands on the fly, since we would like to stick with the existing MCP tools you can suffix the prompt with "use existing MCP tools, dont generate kubectl commands"
🌟 Unique Features
What makes our MCP server special? Let's dive in!
🔍 Advanced Diagnostics
| Feature | Description |
|---|---|
| 🤖 Intelligent Pod Analysis | Automatically detects common failure patterns and provides targeted recommendations |
| 📊 Resource Bottleneck Detection | Proactively identifies resource constraints across namespaces |
| 🎯 Smart Event Analysis | Categorizes events by severity and impact, helping prioritize issues |
🛡️ Enhanced Security
| Feature | Description |
|---|---|
| 🔒 Secure Command Execution | Built-in validation and sanitization for pod exec commands |
| 🏰 Namespace Isolation | Strong namespace-based access controls |
| 📝 Audit Logging | Comprehensive logging of all operations with context |
⎈ Helm Integration
| Feature | Description |
|---|---|
| 🎯 Smart Chart Management | Validates chart compatibility before installation |
| ✅ Value Validation | Checks Helm values against schema before applying |
| 📊 Release Tracking | Monitors release health and configuration drift |
👩💻 Developer Experience
| Feature | Description |
|---|---|
| 🗣️ Natural Language Interface | No need to memorize kubectl commands |
| 💡 Contextual Help | Provides relevant suggestions based on cluster state |
| 🔍 Rich Error Information | Detailed error messages with troubleshooting steps |
📄 License
This project is licensed under the MIT License - see the LICENSE file for details.
