MCP Kubernetes Server

🚀 Kubernetes MCP Server

A lightweight MCP server that combines natural language processing and API access to Kubernetes clusters, integrating kubectl commands with the Kubernetes Python client.

Kubernetes MCP Server

✨ Features

• Natural Language Interface: Transforms plain English queries into kubectl commands.
  • Lists pods and deployments in all namespaces.
  • Falls back to a generic resource list for unsupported queries.
• Full CRUD Operations:
  • 🆕 Creates/deletes namespaces, pods, and deployments via API endpoints.
  • 🔍 Checks cluster resources.
  • ✏️ Modifies labels, annotations, and deployment configurations.
  • 🗑️ Enables graceful deletion with an idle period.
  • 📊 Adjusts deployment scale.
• Dual Execution Modes:
  • Integrates with kubectl commands.
  • Utilizes the official Kubernetes Python client (SDK).
• Advanced Features:
  • Namespace validation (DNS - 1123 compliant).
  • Label filtering.
  • Idle period control.
  • Automatic command fallback.
  • Resource management (CPU, memory).
  • Environment variable configuration.

📦 Installation

Prerequisites

• Python 3.11+
• Access to an available Kubernetes cluster
• Locally configured kubectl
• Installed UV

# Clone the repository
git clone https://github.com/ductnn/mcp-kubernetes-server.git 
cd mcp-kubernetes-server

# Create a virtual environment
uv venv .venv

# Activate (Unix)
source .venv/bin/activate

# Install dependencies
uv pip install -r requirements.txt

🚀 Quick Start

Natural Language Processing

The server supports basic natural language queries to list resources:

# List all pods
result = nl_processor.process("Show me all pods")

# List all deployments
result = nl_processor.process("Show me all deployments")

# Query with a namespace
result = nl_processor.process("Show me all resources", "kube-system")

For more complex operations, use the dedicated API endpoints:

# Create a pod
pod_service.create_pod(
    name="my-pod",
    namespace="default",
    image="nginx:latest",
    labels={"app": "my-app"}
)

# Create a deployment
deployment_service.create_deployment(
    name="my-deployment",
    namespace="default",
    image="nginx:latest",
    replicas=3
)

# Delete a namespace
namespace_service.delete("my-namespace", force=True)

API Endpoints

The server provides RESTful endpoints for all operations:

• /api/pods: Lists all pods.
• /api/deployments: Lists all deployments.
• /api/services: Lists all services.
• /api/namespaces: Lists all namespaces.
• /api/resources: Lists all resources.

🤖 Integration with AI Assistant

Configure as follows:

# Configure the path
export PATH="/path/to/uv:$PATH"

💻 Usage Examples

Basic Usage

# List all pods
result = nl_processor.process("Show me all pods")

# List all deployments
result = nl_processor.process("Show me all deployments")

# Query with a namespace
result = nl_processor.process("Show me all resources", "kube-system")

Advanced Usage

# Create a pod
pod_service.create_pod(
    name="my-pod",
    namespace="default",
    image="nginx:latest",
    labels={"app": "my-app"}
)

# Create a deployment
deployment_service.create_deployment(
    name="my-deployment",
    namespace="default",
    image="nginx:latest",
    replicas=3
)

# Delete a namespace
namespace_service.delete("my-namespace", force=True)

📜 License

This project is licensed under the MIT License.