devops-engineer_skill

Overview

This skill helps design and standardize Universal CI/CD, Multi-Cloud infrastructure, and Observability systems for production-grade software delivery. It codifies principles like Infrastructure as Code, security-first operations, and automated observability to reduce manual toil and increase reliability. Use it to choose cloud providers, IaC and CI/CD tools, and to enforce operational standards across teams.

How this skill works

The skill runs a decision engine that analyzes your stack context and maps requirements to recommended providers and tools (cloud, IaC, CI/CD, code normalization). It references curated comparison guidance to select best-fit options and then loads targeted reference material for implementation patterns, templates, and controls. It also enforces security and observability guardrails like least-privilege IAM, secret management, and mandatory logs/metrics/traces.

When to use it

• Designing a new multi-cloud or cloud-agnostic platform architecture
• Standardizing CI/CD pipelines across teams or repos
• Selecting IaC tools for provisioning and drift control
• Implementing observability and security baseline for production systems
• Migrating workloads between cloud providers or to managed platforms

Best practices

• Automate everything: script repeated tasks and avoid manual click-ops
• Declare infrastructure as code and store remote state securely
• Shift security left: use Vault/KMS for secrets and enforce least privilege
• Mandate observability: collect logs, metrics, and distributed traces
• Choose tools by use case: Terraform for multi-cloud, Pulumi for logic-heavy infra

Example use cases

• Create a GitHub Actions pipeline template that builds, scans (SAST/DAST), and deploys to EKS with Terraform provisioning
• Standardize IaC across teams using Terraform remote state and policy-as-code checks
• Design a multi-cloud data pipeline choosing GCP for ML workloads and AWS for enterprise services
• Implement observability with Prometheus + Grafana and distributed tracing for microservices
• Migrate legacy Jenkins jobs to cloud-native CI (GitLab CI or GitHub Actions) with repeatable pipeline templates