​

MCP Examples

​

Quick Start Examples

​

1. Basic Tool Execution

"Can you check the health of our Kubernetes cluster?"

• kubectl get nodes - Check node health
• kubectl get pods --all-namespaces - List all pods
• kubectl top nodes - Check resource usage

​

2. Execute Custom Script

"Run a Python script that analyzes our application logs"

• Create a Python container
• Execute your analysis script
• Return results with live streaming

​

3. Multi-Step Workflow

"Deploy version 2.1.0 to staging, run tests, and promote to production if successful"

• Build Docker image
• Deploy to staging
• Run integration tests
• Promote to production if tests pass

​

AI Assistant Integrations

​

Claude Desktop

{
  "mcpServers": {
    "kubiya": {
      "command": "kubiya",
      "args": ["mcp", "serve"],
      "env": {
        "KUBIYA_API_KEY": "your-api-key"
      }
    }
  }
}

"List all runners and check their health status"
"Execute kubectl get pods in production namespace"
"Check disk space on all servers"

"Create a backup of our PostgreSQL database"
"Deploy the latest version of our API service"
"Run our nightly data processing job"

​

Cursor IDE

{
  "mcp.servers": {
    "kubiya": {
      "command": "kubiya",
      "args": ["mcp", "serve"],
      "env": {
        "KUBIYA_API_KEY": "your-api-key"
      }
    }
  }
}

"Use Kubiya to run our test suite on the dev cluster"
"Deploy this branch to staging for testing"
"Check the logs for any errors in the last hour"

"Run security scans on our codebase"
"Check code coverage for our tests"
"Analyze performance metrics for our API"

​

Real-World Use Cases

​

1. DevOps Assistant

1. Checks application metrics using monitoring tools
2. Analyzes recent logs for errors
3. Examines resource usage (CPU, memory, disk)
4. Identifies bottlenecks
5. Suggests optimizations

• execute_tool with kubectl commands
• execute_tool with custom monitoring scripts
• search_kb for troubleshooting guides

​

2. Database Operations

# Create timestamped backup
TIMESTAMP=$(date +%Y%m%d_%H%M%S)
pg_dump -h prod-db -U backup_user myapp > backup_${TIMESTAMP}.sql

# Compress backup
gzip backup_${TIMESTAMP}.sql

# Upload to S3
aws s3 cp backup_${TIMESTAMP}.sql.gz s3://db-backups/myapp/

​

3. Security Scanning

1. Scans container images for vulnerabilities
2. Checks Kubernetes configurations for security issues
3. Reviews access controls and permissions
4. Validates compliance with security policies
5. Generates comprehensive security report

​

4. Data Processing Pipeline

1. Extracts data from multiple sources (databases, APIs)
2. Runs ETL transformations using pandas/numpy
3. Validates data quality
4. Updates data warehouse
5. Refreshes BI dashboards

​

Advanced Examples

​

1. Multi-Cloud Deployment

• AWS ECS deployment
• GCP Cloud Run deployment
• Load balancer configuration
• Health checks for both environments

​

2. Incident Response

1. Gathers system metrics and logs
2. Identifies root cause
3. Implements immediate fixes
4. Scales resources if needed
5. Documents incident timeline

​

3. CI/CD Pipeline

• GitHub Actions workflow
• Docker build and push
• Kubernetes deployment
• Testing and validation
• Rollback mechanisms

​

Integration Patterns

​

1. Tool Chaining

"First check the database connections, then restart any failed services, and finally verify everything is working"

​

2. Conditional Execution

"If the staging tests pass, deploy to production, otherwise send me the test results"

​

3. Parallel Processing

"Check the health of all our services and generate a status report"

​

Configuration Examples

​

Basic Configuration

{
  "enable_runners": true,
  "allow_platform_apis": false,
  "enable_opa_policies": false,
  "verbose_logging": false
}

​

Enterprise Configuration

{
  "enable_runners": true,
  "allow_platform_apis": true,
  "enable_opa_policies": true,
  "verbose_logging": true,
  "whitelisted_tools": [
    {
      "name": "kubectl",
      "description": "Kubernetes CLI tool",
      "type": "docker",
      "image": "kubiya/kubectl-light:latest"
    }
  ]
}

​

Error Handling

​

Graceful Failure

"The kubectl command failed because the cluster is unreachable. 
Would you like me to try connecting to the backup cluster?"

​

Retry Logic

"Retrying connection to database... (attempt 2/3)"

​

Monitoring and Logging

​

Execution Tracking

• Execution time
• Exit codes
• Resource usage
• Error messages

​

Debug Mode

kubiya mcp serve --verbose

​

Best Practices

​

1. Security First

• Always use appropriate access controls
• Enable OPA policies for production
• Regularly rotate API keys
• Monitor tool execution logs

​

2. Resource Management

• Use appropriate runners for different workloads
• Monitor resource usage
• Set timeouts for long-running tasks

​

3. Error Handling

• Implement proper error handling
• Use retries for transient failures
• Provide meaningful error messages

​

Next Steps

• Explore the Tools Reference
• Setup Authentication & Policies
• Build Full Stack Agents