Docker Management Skill

Version: 1.0.0 Purpose: Docker container and compose management

Triggers

Integrated MCP Tools

Container Operations

Compose Operations

Workflow: Container Debugging

Phase 1: Assessment

Step 1.1: List Containers

Use docker_ps with:
- all: true (include stopped)

Step 1.2: Check Status

Look for:

• Exit codes
• Restart counts
• Health status

Phase 2: Investigation

Step 2.1: View Logs

Use docker_logs with:
- container: Container name/ID
- tail: 100 (last N lines)
- timestamps: true

Step 2.2: Inspect Configuration

Use docker_inspect to check:
- Environment variables
- Mount points
- Network settings
- Health check config

Step 2.3: Resource Usage

Use docker_stats to monitor:
- CPU percentage
- Memory usage
- Network I/O
- Block I/O

Phase 3: Resolution

Step 3.1: Restart Container

Use docker_restart for temporary fix

Step 3.2: Execute Commands

Use docker_exec to run diagnostics inside container:
- command: "sh -c 'ps aux'"
- command: "cat /var/log/app.log"

Workflow: Docker Compose

Step 1: Check Status

Use compose_ps to see all services

Step 2: View Logs

Use compose_logs with:
- service: Specific service name
- tail: 50

Step 3: Service Management

Use compose_up to start services
Use compose_down to stop and clean up

Common Issues

Best Practices

✅ GOOD:

• Check logs before restarting
• Use health checks
• Set resource limits
• Use named volumes

❌ BAD:

• Restart without investigation
• Run as root unnecessarily
• Store secrets in images
• Use latest tag in production

Checklist

• Docker daemon running
• Container status checked
• Logs reviewed
• Resource usage normal
• Network connectivity OK