Cursor Integration

Coming Soon — V2.00: MCP integration is coming as part of Kubeshark V2.00. Read the announcement.

Cursor is an AI-powered code editor with native MCP support. Connect Kubeshark to query and analyze Kubernetes network traffic directly from your IDE using natural language.

Quick Setup

Option 1: UI Configuration

  1. Open Cursor Settings (File → Preferences → Cursor Settings)
  2. Select MCP from the sidebar
  3. Click Add new global MCP server
  4. Enter the configuration

Option 2: Configuration File

Cursor supports two configuration locations:

LocationFileScope
Global~/.cursor/mcp.jsonAll projects
Project.cursor/mcp.jsonCurrent project only

Configuration

URL Mode

Connect to an existing Kubeshark deployment:

{
  "mcpServers": {
    "kubeshark": {
      "command": "kubeshark",
      "args": ["mcp", "--url", "https://kubeshark.example.com"]
    }
  }
}

Proxy Mode

Let the CLI proxy into your cluster via kubectl:

{
  "mcpServers": {
    "kubeshark": {
      "command": "kubeshark",
      "args": ["mcp", "--kubeconfig", "/path/to/.kube/config"]
    }
  }
}

To enable cluster management operations (start/stop Kubeshark):

{
  "mcpServers": {
    "kubeshark": {
      "command": "kubeshark",
      "args": ["mcp", "--allow-destructive", "--kubeconfig", "/path/to/.kube/config"]
    }
  }
}

Project-Specific Configuration

For team sharing, create .cursor/mcp.json in your project root:

{
  "mcpServers": {
    "kubeshark": {
      "command": "kubeshark",
      "args": ["mcp", "--url", "https://kubeshark.example.com"]
    }
  }
}

Team sharing: Commit .cursor/mcp.json to version control so your team automatically gets Kubeshark integration.

Managing MCP Servers

Cursor provides CLI commands for MCP management:

# List configured MCP servers
/mcp list

# Enable a server
/mcp enable kubeshark

# Disable a server
/mcp disable kubeshark

Example Prompts

Once connected, ask Cursor’s AI:

“What services are running in my Kubernetes cluster?”

“Show me any HTTP 500 errors in the last hour.”

“Which pods are communicating with the payment service?”

“Find the API calls that took longer than 500ms.”

“What’s causing the latency between order-service and inventory-db?”

Combining Network Data with Code

Since Cursor has access to both your codebase and Kubeshark’s network data, you can ask questions that span both:

“The payment-service is returning 503 errors. Find where the timeout is configured in the code.”

“Show me the API calls to stripe-gateway and find the retry logic in the codebase.”

“There’s a slow query to postgres. Find the code that makes this query and suggest optimizations.”

Troubleshooting

MCP server not appearing

  • Verify the Kubeshark binary is in your PATH or use an absolute path
  • Restart Cursor after adding the configuration
  • Check the MCP panel in Cursor Settings for error messages

Connection errors

# Test the binary directly
kubeshark mcp --list-tools --url https://kubeshark.example.com

# If using proxy mode, verify kubectl access
kubectl get pods -l app=kubeshark-hub

Server not responding

  • Ensure Kubeshark Hub is running and accessible
  • If using --url mode, verify the URL is reachable
  • Check that L7 dissection is enabled if you need API-level data

What’s Next