QUICKSTART.md•3.22 kB
# Quick Start Guide: kubectl-mcp-tool
This guide will help you quickly set up kubectl-mcp-tool to work with AI assistants like Cursor, Claude, and WindSurf.
## 1. One-Step Installation
The simplest way to install and configure kubectl-mcp-tool is to use our installation script:
```bash
# Clone the repository
git clone https://github.com/rohitg00/kubectl-mcp-server.git
cd kubectl-mcp-server
# Run the installation script
bash install.sh
```
This script will:
- Install the required dependencies
- Set up configurations for Cursor, Claude, and WindSurf
- Configure environment variables correctly
- Test your Kubernetes connection
## 2. Manual Installation
If you prefer to install manually:
```bash
# Install the package
pip install mcp>=1.5.0
pip install -e .
# Create config directories
mkdir -p ~/.cursor
mkdir -p ~/.config/claude
mkdir -p ~/.config/windsurf
```
## 3. Configuration for AI Assistants
### Cursor
Create or edit `~/.cursor/mcp.json`:
```json
{
"mcpServers": {
"kubernetes": {
"command": "python",
"args": ["-m", "kubectl_mcp_tool.minimal_wrapper"],
"env": {
"KUBECONFIG": "/path/to/your/.kube/config",
"PATH": "/usr/local/bin:/usr/bin:/bin:/usr/sbin:/sbin:/opt/homebrew/bin"
}
}
}
}
```
### Claude Desktop
Create or edit `~/.config/claude/mcp.json`:
```json
{
"mcpServers": {
"kubernetes": {
"command": "python",
"args": ["-m", "kubectl_mcp_tool.minimal_wrapper"],
"env": {
"KUBECONFIG": "/path/to/your/.kube/config"
}
}
}
}
```
### WindSurf
Create or edit `~/.config/windsurf/mcp.json`:
```json
{
"mcpServers": {
"kubernetes": {
"command": "python",
"args": ["-m", "kubectl_mcp_tool.minimal_wrapper"],
"env": {
"KUBECONFIG": "/path/to/your/.kube/config"
}
}
}
}
```
> **Important**: Replace `/path/to/your/.kube/config` with the actual path to your kubeconfig file (usually `~/.kube/config`).
## 4. Testing Your Installation
To verify your installation is working:
1. Test the command line:
```bash
kubectl-mcp --help
```
2. Test the MCP server directly:
```bash
python -m kubectl_mcp_tool.minimal_wrapper
```
It should start and wait for connections with no errors.
3. Open your AI assistant (Cursor, Claude, or WindSurf) and ask a Kubernetes-related question:
- "List all pods in the default namespace"
- "What deployments are running in my cluster?"
- "Show me the services in the kube-system namespace"
## 5. Troubleshooting
If you encounter issues:
1. Check the logs:
```bash
# For Cursor
cat ~/.cursor/logs/kubectl_mcp_debug.log
```
2. Run the tool directly to see error messages:
```bash
python -m kubectl_mcp_tool.minimal_wrapper
```
3. Verify kubectl is working from command line:
```bash
kubectl get pods
```
4. Check if your configuration file has the correct paths:
```bash
cat ~/.cursor/mcp.json
```
5. Try reinstalling the package:
```bash
pip uninstall -y kubectl-mcp-tool
pip install -e .
```
For more details, see the [Installation Guide](./INSTALLATION.md) and [Troubleshooting](./INSTALLATION.md#troubleshooting) sections.