afrexai-devops-engine_skill

Overview

This skill is a complete DevOps and Platform Engineering engine for teams building, deploying, and operating production software across any cloud. It bundles opinionated patterns for branching, CI/CD pipelines, container strategy, infrastructure-as-code, and Kubernetes operations. The goal is reproducible, observable, and secure delivery pipelines that scale with team size and compliance requirements.

How this skill works

The engine inspects repository layout and applies a recommended branch and commit strategy, then generates a reusable CI/CD pipeline template that enforces build-once deploy-everywhere and fail-fast principles. It provides Dockerfile patterns, caching strategies, and security scanning steps for images. For infrastructure it prescribes IaC layouts, remote state safety rules, and an environment promotion workflow; for Kubernetes it supplies production-ready manifests, autoscaling, and Helm chart checklists.

When to use it

• Setting up a new microservice or monorepo with consistent CI/CD and branch controls
• Migrating from ad-hoc build scripts to reproducible, hermetic pipelines and immutable artifacts
• Standardizing container builds, image scanning, and local developer compose setups
• Implementing multi-environment IaC with remote state and safe promotion gates
• Hardening Kubernetes deployments: probes, resource limits, autoscaling, and rollout strategies
• Preparing for audits or regulated releases with signed tags and stricter branch protection

Best practices

• Adopt a branch strategy matched to team size and release cadence; protect main with mandatory checks and PRs
• Build once and tag artifacts with the Git SHA; promote the same artifact through dev→staging→prod
• Fail fast: run lint/typecheck/security in parallel before slower tests; enforce coverage thresholds
• Use multi-stage Docker builds, minimal base images, and image scanners like Trivy in CI
• Keep Terraform state remote with locking, always plan before apply, and protect critical resources from destroy
• Define Kubernetes readiness/liveness probes, requests/limits, PDBs, and run non-root containers

Example use cases

• Generate a reusable GitHub Actions workflow that builds, scans, and deploys canary releases to production
• Create a Terraform project structure with remote S3/GSM state and review-in-PR promotion to environments
• Replace manual docker builds with multi-stage Dockerfiles and integrate Trivy vulnerability gating in CI
• Ship a Helm chart that includes network policies, external secrets integration, and CI linting/templates in pipeline
• Operate a Kubernetes service with HPA, rolling updates, topology spread, and controlled rollback windows