devops-expert_skill

Overview

This skill turns the agent into a practical DevOps and cloud-native expert who designs, implements, and operates modern CI/CD pipelines, containerized workloads, Kubernetes deployments, and infrastructure automation. It focuses on actionable guidance for production-ready systems, reliability, security, and cost-efficient cloud operations. Use it to accelerate delivery, improve observability, and automate repetitive operational work.

How this skill works

The skill inspects your application architecture, CI/CD needs, and target cloud platform to propose concrete pipelines, infrastructure-as-code, and deployment manifests. It generates recommended configurations for GitHub Actions/GitLab CI/Jenkins, Docker and Kubernetes manifests (including Helm charts), Terraform modules, and monitoring stacks like Prometheus/Grafana. It also provides security hardening, backup/disaster plans, and operational runbooks tailored to your environment.

When to use it

• Setting up or improving CI/CD pipelines for automated build-test-deploy
• Containerizing applications and deploying to Kubernetes or managed clusters
• Designing cloud-native architectures and multi-environment strategies
• Automating cloud infrastructure with Terraform, Ansible, or ARM/CloudFormation
• Implementing monitoring, logging, and distributed tracing for production systems

Best practices

• Adopt least-privilege IAM and secrets management (Vault, AWS Secrets Manager)
• Use multi-stage Docker builds, image scanning, and minimal base images
• Treat infrastructure as code with modular, reusable Terraform modules and state management
• Design observability from day one: metrics, logs, and tracing with alerting and runbooks
• Implement progressive deployment strategies (blue/green, canary) and automated rollbacks

Example use cases

• Create a GitHub Actions pipeline that runs tests, builds Docker images, pushes to registry, and deploys to Kubernetes
• Design a highly available microservices platform on EKS/GKE/AKS with autoscaling, service mesh option, and ingress routing
• Write Terraform to provision VPC, EKS cluster, RDS, and S3 with tagging and cost controls
• Assemble a monitoring stack: Prometheus for metrics, Grafana dashboards, Loki/ELK for logs, and Jaeger for tracing
• Audit an existing CI/CD pipeline for security gaps, flaky deployments, and performance bottlenecks