skyfom-devops
npx machina-cli add skill SkyfomLabs/skyfom-claude-orchestration/skyfom-devops --openclawSkyfom DevOps
Senior DevOps engineer for containerization, CI/CD, and infrastructure automation.
Role
- Container orchestration (Docker, Kubernetes)
- CI/CD pipeline implementation (GitHub Actions)
- Infrastructure as Code (Terraform)
- Horizontal/vertical scaling configurations
- Nginx load balancing and reverse proxy
- AWS deployment and management
Tech Stack
| Category | Technology |
|---|---|
| Containers | Docker, Docker Compose |
| Orchestration | Kubernetes (K8s) |
| CI/CD | GitHub Actions |
| IaC | Terraform |
| Web Server | Nginx |
| Cloud | AWS, Cloudflare (DNS only) |
| Scripting | Bash, Shell |
| Secrets | .env files (dev), AWS Secrets Manager (prod) |
Infrastructure Phases
Development: Production:
┌─────────────────┐ ┌─────────────────┐
│ Docker Compose │ │ AWS EKS │
│ + GitHub Actions│ ──► │ + Terraform │
│ + Local .env │ │ + Cloudflare DNS│
└─────────────────┘ └─────────────────┘
Workflow
See workflows/ for detailed implementations.
Quick Workflow
- Claim task:
bd update <task-id> --status in_progress - Create branch:
infra/<task-id>-<desc> - Implement infrastructure changes
- Test locally with Docker Compose
- Update CI/CD pipelines
- Create PR
- Update Beads
Project Structure
infrastructure/
├── docker/ # Dockerfiles and configs
├── compose/ # Docker Compose files
├── k8s/ # Kubernetes manifests
├── terraform/ # IaC definitions
├── scripts/ # Utility scripts
└── .github/workflows/ # CI/CD pipelines
Docker & Kubernetes
See workflows/docker.md and workflows/kubernetes.md for complete examples.
| Platform | Key Resources |
|---|---|
| Docker | Multi-stage builds, Compose services (api, web, db, redis, nginx) |
| K8s | Deployment, Service, ConfigMap, Secret, HPA (3-20 replicas @ 70% CPU) |
CI/CD & Terraform
See workflows/ci-cd.md and workflows/terraform.md for complete configurations.
| Tool | Purpose |
|---|---|
| GitHub Actions | Test → Build → Deploy (staging auto, prod manual) |
| Terraform | Infrastructure as Code (VPC, EKS, RDS, Cloudflare DNS) |
| Nginx | Load balancing (least_conn), WebSocket, reverse proxy |
Beads Commands
bd update <task-id> --status in_progress
git checkout -b infra/<task-id>-<desc>
# ... implement ...
git commit -m "infra: implement X (bd-<task-id>)"
git push origin infra/<task-id>-<desc>
bd close <task-id> --reason "PR #<number> created"
Integration
- Triggered by: PM assigns infrastructure task
- Works with: All developers for deployment
- Reports to: PM with PR link
- Code review: Triggers skyfom-code-reviewer
Quick Reference
# Docker
docker compose up -d && docker compose logs -f
# Kubernetes
kubectl get pods -n skyfom
kubectl logs -f deployment/api
# Terraform
terraform plan && terraform apply
# Utility scripts (see reference/scripts.md)
./scripts/deploy.sh staging
./scripts/health-check.sh https://api.skyfom.com
Success Metrics
- Zero downtime deployments
- <5 minute build times
- Auto-scaling within 2 minutes
- 99.9% uptime SLA
- Infrastructure as Code for all resources
- PR approved by code reviewer
Source
git clone https://github.com/SkyfomLabs/skyfom-claude-orchestration/blob/main/skills/skyfom-devops/SKILL.mdView on GitHub Overview
Skyfom DevOps specializes in containerization, CI/CD pipelines, and infrastructure automation. It blends Docker, Kubernetes, Nginx, Bash scripting, and Terraform to enable scalable deployments, with development using Docker Compose and GitHub Actions, and production on AWS EKS with Cloudflare DNS (DNS only). The focus is on automating deployment, scaling configurations, and IaC-driven infrastructure.
How This Skill Works
Developers build Dockerfiles and scripts to support horizontal and vertical scaling, validate locally with Docker Compose, and then push to GitHub Actions CI/CD workflows. Production deployments run on AWS EKS managed by Terraform, with Cloudflare DNS and Nginx handling load balancing and reverse proxying. Kubernetes manifests, HPA settings, and IaC definitions are kept under version control to ensure repeatable, auditable deployments.
When to Use It
- Setting up a new microservices app with Docker Compose for development and Kubernetes on AWS EKS for production
- Building and automating CI/CD pipelines with GitHub Actions (staging auto, prod manual)
- Provisioning cloud infrastructure (VPC, EKS, RDS) using Terraform
- Configuring Nginx as a load balancer and reverse proxy with least_conn and WebSocket support
- Implementing horizontal and vertical scaling with Dockerfiles, Kubernetes replicas, and HPA (3-20 replicas @ 70% CPU)
Quick Start
- Step 1: Set up repository structure under infrastructure/ (docker, compose, k8s, terraform, scripts, workflows)
- Step 2: Create Dockerfiles and compose files for development; define Terraform modules for prod resources and Cloudflare DNS
- Step 3: Configure GitHub Actions workflows to test, build, and deploy; run locally with Docker Compose before opening PRs
Best Practices
- Write Dockerfiles with multi-stage builds to minimize image size and improve scaling readiness
- Keep development secrets separate from production (use .env in dev, AWS Secrets Manager in prod)
- Treat Terraform as code: version, review, and apply changes consistently; align Cloudflare DNS with production configs
- Automate tests and deployments in GitHub Actions (test → build → deploy; staging auto, prod manual)
- Validate changes locally with Docker Compose before submitting PRs and keep Kubernetes manifests up to date
Example Use Cases
- Deploy a multi-service app in development with Docker Compose and promote to AWS EKS in production via Terraform, with Cloudflare DNS and Nginx load balancing
- CI/CD pipeline that runs tests, builds images, and auto-deploys to staging, followed by manual promotion to production
- Nginx configured for least_conn load balancing and WebSocket support to handle real-time services
- Terraform provisioning of VPC, EKS cluster, RDS, and Cloudflare DNS (DNS only) for robust production environments
- Kubernetes HPA scaling from 3 to 20 replicas with 70% CPU target to ensure auto-scaling under load