An Internal Developer Platform for Self-Service Cloud-Native Application Delivery on AWS
Status: Phase 1 of 18 (architecture and repository foundation). Nothing in this repository has been deployed yet. Sections are marked with the phase that delivers them; this README will be updated as the platform is built. Badges are added only once the workflows behind them exist.
CloudForge lets a developer request a standardised microservice and get a repo, CI/CD, Helm config, GitOps registration, observability and security guardrails automatically, using Terraform/Terragrunt, GitHub Actions, ArgoCD and Amazon EKS. Principle: golden paths over custom snowflakes.
See docs/architecture/overview.md (system context, component, delivery-flow and tracing diagrams) and technology decisions.
Every component has a documented reason, alternatives, security/cost/operational implications and failure modes: see the ADRs.
| Area | Status |
|---|---|
| Terraform/Terragrunt AWS foundation (VPC, EKS, RDS, ECR, KMS, IAM) | Phase 2-3 |
| Kubernetes platform (RBAC, NetworkPolicies, Kyverno, ESO) | Phase 4, 10 |
| user/order/payment FastAPI services | Phase 5 |
| GitHub Actions CI with OIDC and security gates | Phase 6-7 |
| ArgoCD app-of-apps GitOps, drift and rollback demo | Phase 8 |
| Prometheus, Grafana, Loki, OpenTelemetry, SLOs, alerts | Phase 9 |
| Autoscaling (HPA + Karpenter), FinOps | Phase 11-12 |
| DR, chaos experiments, runbooks | Phase 13-14 |
| cloudforge-api and CLI | Phase 15-16 |
cloudforge/
├── infrastructure/
│ ├── terraform/modules/ # vpc eks rds ecr iam kms s3 route53 security-groups monitoring
│ └── terragrunt/environments/{dev,staging,production}
├── platform/ # argocd, kyverno, external-secrets, karpenter, observability, network-policies, rbac
├── gitops/{dev,staging,production} # desired state (own repo from Phase 8)
├── helm/service-template/ # shared chart
├── services/ # user-service, order-service, payment-service
├── platform-api/ # cloudforge-api (FastAPI)
├── cli/ # cloudforge CLI (Typer)
├── templates/service-python/ # golden-path service template
├── local/ # docker-compose, kind
├── runbooks/ # operational runbooks
├── docs/ # architecture, adr, security, dr, chaos, roadmap, demo
├── tests/ # policy, pipeline, integration
├── scripts/ # destroy.sh and helpers
└── .github/ # workflows, CODEOWNERS
Docker, Python 3.12+, kubectl, helm, kind, terraform, terragrunt, make, AWS account (from Phase 2). See local development.
Intentionally no figures yet. See cost control for expensive resources; measured numbers arrive in Phase 12. Run make destroy ENV=<env> when finished.
Terraform, Kubernetes, GitHub Actions, ArgoCD, Observability, Autoscaling, Disaster Recovery, Chaos Engineering, Troubleshooting, Runbooks, Demo, Screenshots, Future Improvements, Lessons Learned.
Add your name and links here.
MIT