Backstage, ArgoCD, Kargo로 구성된 검증된 스택을 사용하여 Amazon EKS 기반의 Internal Developer Platform(IDP)을 설계하고 구축합니다.
- GitHub 스타
- 97
- 파일
- 12
- 마지막 커밋
- 2026년 11월 12일
- 라이선스
- MIT-0
- 허용된 도구
- —
eks-platform-engineering skill은 무엇을 하나요?
Backstage, ArgoCD, Kargo로 구성된 검증된 스택을 사용하여 Amazon EKS 기반의 Internal Developer Platform(IDP)을 설계하고 구축합니다.
skill의 SKILL.md에서 생성된 요약입니다.
eks-platform-engineering skill 정보
이 SKILL의 주요 목표는 Amazon EKS 상에 Internal Developer Platform을 설계하고 구축하는 것입니다. 표준화된 golden path를 통해 개발자의 셀프 서비스 기능을 구현하려는 플랫폼 엔지니어를 위한 최적화된 아키텍처 청사진을 제공합니다. 개별 클러스터를 관리하는 대신, 중앙 제어 평면이 GitOps 원칙에 따라 워크로드 클러스터를 관리하는 허브 앤 스포크(hub-and-spoke) 토폴로지를 구축하는 방법을 학습합니다.
이 SKILL은 개발자 포털용 Backstage, GitOps 배포용 ArgoCD, 다단계 아티팩트 프로모션용 Kargo를 포함한 특정 도구 스택의 통합 과정을 안내합니다. 또한 ACK와 kro를 사용하여 복잡한 AWS 리소스를 추상화하는 방법과 Argo Rollouts를 사용하여 점진적 배포를 구현하는 방법을 다룹니다. 개발자는 Kiro를 사용하여 플랫폼 템플릿 및 매니페스트 생성을 지원받을 수 있습니다. 이러한 접근 방식을 통해 애플리케이션 팀은 Kubernetes나 AWS API에 직접 접근할 필요 없이 간단한 템플릿만으로 환경과 리소스를 프로비저닝할 수 있습니다.
agent가 수행할 수 있는 작업
- 1개발자 셀프 서비스 포털 구축개발자가 소프트웨어 템플릿과 서비스 카탈로그에 접근할 수 있는 중앙 집중식 엔트리 포인트를 설정합니다.
- 2AWS 리소스 프로비저닝 자동화개발자가 Kubernetes 매니페스트를 통해 S3 또는 DynamoDB와 같은 클라우드 리소스를 요청할 수 있도록 합니다.
- 3표준화된 개발자 golden path 생성새로운 애플리케이션이 보안 및 운영 가드레일을 상속받을 수 있도록 표준화된 경로를 정의합니다.
skill의 SKILL.md를 기반으로 작성되었습니다. 모든 사용 사례는 해당 문서의 인용구와 대조 확인되었습니다.
eks-platform-engineering skill 설치 방법
각 명령은 skill 폴더를 클라이언트의 skills 디렉터리에 복사합니다. 클라이언트를 재시작하거나 새 세션을 시작하여 로드하세요.Claude Code에 eks-platform-engineering skill 설치
~/.claude/skills/eks-platform-engineering/SKILL.md
$ npx degit aws-samples/sample-apex-skills/skills/eks-platform-engineering ~/.claude/skills/eks-platform-engineeringClaude Code (project)에 eks-platform-engineering skill 설치
.claude/skills/eks-platform-engineering/SKILL.md
$ npx degit aws-samples/sample-apex-skills/skills/eks-platform-engineering .claude/skills/eks-platform-engineeringCodex에 eks-platform-engineering skill 설치
~/.agents/skills/eks-platform-engineering/SKILL.md
$ npx degit aws-samples/sample-apex-skills/skills/eks-platform-engineering ~/.agents/skills/eks-platform-engineeringGemini CLI에 eks-platform-engineering skill 설치
~/.gemini/skills/eks-platform-engineering/SKILL.md
$ npx degit aws-samples/sample-apex-skills/skills/eks-platform-engineering ~/.gemini/skills/eks-platform-engineeringClaude.ai에 eks-platform-engineering skill 설치
npx degit aws-samples/sample-apex-skills/skills/eks-platform-engineering eks-platform-engineering && zip -r eks-platform-engineering.zip eks-platform-engineering # claude.ai → Customize → Skills → + → Create skill → Upload a skill → eks-platform-engineering.zip
자동 검사
2026년 11월 14일에 실행됨전체 보고서 →자동 검사를 통과했습니다. 보안 감사는 아닙니다. 에이전트가 실행하기 전에 스크립트를 읽어 보세요. 검사 방식
- 소스 저장소 연결 및 접근 가능aws-samples/sample-apex-skills/skills/eks-platform-engineering is public and reachablegithub.com/aws-samples/sample-apex-skills/tree/HEAD/skills/e ↗통과
- 오픈 소스 라이선스 (OSI)MIT-0spdx.org/licenses/MIT-0.html ↗통과
- 최근 1년 내 커밋 또는 릴리스Last commit 2026-11-12github.com/aws-samples/sample-apex-skills/commits ↗통과
- 스크립트에 다운로드 후 실행이나 자격 증명 탈취가 없음No scripts in the skill folder통과
- SKILL.md에 숨겨진 지침 없음No override phrases, safeguard switches, credential requests or hidden charactersgithub.com/aws-samples/sample-apex-skills/blob/HEAD/skills/e ↗통과
- 사전 허용 도구가 제한됨No pre-approved tools; the agent asks before using onegithub.com/aws-samples/sample-apex-skills/blob/HEAD/skills/e ↗통과
- SKILL.md가 Agent Skills 형식을 따름Valid name and description; says when to use it; 162 linesgithub.com/aws-samples/sample-apex-skills/blob/HEAD/skills/e ↗통과
SKILL.md
GitHub에서 보기EKS Platform Engineering
Guidance for designing and building an Internal Developer Platform (IDP) on Amazon EKS. This skill is opinionated: it teaches one proven, integrated golden-path stack end to end rather than cataloguing every option. The reference architecture and tool choices below are the recommended default; deviate only with a reason.
This is the "how do I build a platform that other teams self-serve from" skill. For "how do I run a single EKS cluster well" (compute, networking, security, cost, upgrades), use eks-best-practices instead.
When to Use This Skill
Activate when the user wants to:
- Build or design an Internal Developer Platform / developer self-service on EKS
- Stand up or wire together a developer portal (Backstage), GitOps (ArgoCD/Argo Workflows), progressive delivery (Argo Rollouts), or promotion (Kargo)
- Define golden paths — standardized, paved ways for app teams to ship
- Abstract AWS resources behind Kubernetes (ACK, kro) or applications behind a developer-facing model (Backstage templates + kro, or OAM/KubeVela)
- Enable self-service environment or resource provisioning
- Measure platform success (DORA metrics)
- Use GenAI (Kiro) to author platform templates/manifests
- Extend the platform to AI/ML or data-engineering workloads
Don't use this skill for:
- Single-cluster EKS architecture, sizing, cost, or upgrade decisions with no self-service/platform angle →
eks-best-practices - Standalone Terraform/OpenTofu module authoring →
terraform-skill - Discovering what's already running on a cluster →
eks-recon - Generic Kubernetes concepts (Claude knows these)
What Is an Internal Developer Platform
CNCF defines a platform as "an integrated collection of capabilities defined and presented according to the needs of the platform's users." An IDP packages tools, services, and automation so application, ML, and data teams self-serve — provisioning environments, resources, and deployments without tickets — while a platform team owns the paved paths and guardrails.
Why it matters: speed (faster delivery), control (safe, consistent operations), cost (economies of scale), and continuous improvement (shared golden paths). Platform engineering has moved from emerging trend to mainstream practice — Gartner's widely-cited forecast that 80% of large software organizations would stand up platform engineering teams by 2026 reflects how quickly the discipline has been adopted.
The core principle — separation of concerns:
- Platform team defines how: portal templates, the application abstraction (Backstage software templates + kro compositions, or OAM component/trait definitions), ACK controllers, CI/CD scaffolds, promotion stages, guardrails.
- App/ML/data teams choose what: pick a template, fill parameters, push code. They never touch raw Kubernetes or AWS APIs.
See references/idp-architecture.md for the full reference architecture and the IDP value/challenge framework.
The Opinionated Platform Stack
This is the recommended, integration-tested stack. Each layer has one default tool.
| Layer | Tool | Role |
|---|---|---|
| Developer portal | Backstage | Self-service catalog + software templates (scaffolder) — the "front door" |
| Identity / SSO | Keycloak | One login (OIDC/SAML) federated across every platform tool |
| GitOps CD | ArgoCD | Reconciles cluster state from Git; deploys platform add-ons and apps |
| CI / orchestration | Argo Workflows (+ Argo Events) | Container-native build pipelines, webhook-triggered |
| Progressive delivery | Argo Rollouts | Canary/blue-green with functional, performance, and metrics gates + auto-rollback |
| Multi-stage promotion | Kargo | GitOps-native dev→prod promotion of the same artifact |
| AWS resource provisioning | ACK (AWS Controllers for K8s) | AWS resources (S3, DynamoDB, IAM, …) as Kubernetes CRDs |
| Resource composition | kro | Compose many resources into one CRD (e.g. CICDPipeline, EKSCluster) |
| Application abstraction | Backstage templates + kro | Developer-facing app model — a Backstage template gathers parameters and a kro ResourceGraphDefinition renders the underlying workload + dependencies (KubeVela/OAM is a supported alternative — see note) |
| Secrets | External Secrets Operator | Sync secrets from AWS Secrets Manager into clusters |
| Identity for workloads | EKS Pod Identity (recommended for new workloads) / IRSA | Credential-free AWS access from pods |
| Observability / metrics | Amazon Managed Grafana + Prometheus + Apache DevLake | Dashboards, app metrics, DORA metrics |
| GenAI | Kiro | Generate Backstage templates, kro/OAM compositions, and deployment manifests (Kiro is AWS's successor to Amazon Q Developer — see GenAI section) |
Maturity note on
kro: kro (which bundles many resources into one simple custom resource) is newer and less battle-tested than the rest of this stack — as of June 2026 it is at v0.9.2 with av1alpha1API (pre-1.0; the maintainers note breaking changes may still land). It works well and is a Kubernetes SIG subproject, but confirm its current maturity fits your risk tolerance before you standardize on it. If you want a more established tool for the same job — composing AWS and Kubernetes resources behind one API — Crossplane is the proven alternative.Application-abstraction choice: the default above (Backstage templates + kro) keeps the platform on one composition engine. KubeVela / OAM is a fully supported alternative that gives a richer
Applicationmodel (Components + Traits) — choose it if you want that abstraction, but note its current velocity before standardizing on it (see references/application-model-oam.md).
Cluster topology — hub-and-spoke (default): a hub cluster runs the platform control plane (ArgoCD, Backstage, GitLab, Keycloak, Kargo); spoke clusters (dev, prod) run workloads. EKS Auto Mode is the default cluster type so the platform team isn't managing nodes. Infrastructure metadata flows to clusters via the GitOps Bridge pattern. Details: references/idp-architecture.md.
The Golden Paths
A golden path is a paved, opinionated route from intent to running software, with guardrails baked in. The platform ships these as Backstage templates. The four core paths:
- Provision an environment — Backstage template → kro composes VPC + EKS Auto Mode cluster + add-ons + ArgoCD registration. Minutes, not days. → references/infrastructure-abstraction.md
- Provision an AWS resource — Backstage template → ACK manifest committed to Git → ArgoCD syncs → ACK provisions the AWS resource. → references/infrastructure-abstraction.md
- Onboard an application — Backstage template → kro
CICDPipeline→ Argo Workflows CI + ArgoCD CD. Developer contract: provide aDockerfile+ the platform's application manifest (a kro-rendered resource, or an OAMApplicationif you run the KubeVela abstraction); the platform does the rest. → references/developer-portal-backstage.md, references/gitops-delivery.md - Ship safely to production — Argo Rollouts canary with gates in dev; Kargo promotes the validated artifact to prod (auto in dev, manual approval for prod). → references/progressive-delivery.md
The developer contract (what app teams provide vs. what the platform handles) is the design heart of every golden path — see references/golden-paths.md.
Guardrails — making the paved path the safe path
Self-service is only safe if the guardrails are built into the golden path, so they apply automatically and a developer cannot accidentally skip them:
- Policy-as-code — an admission policy engine (Kyverno or OPA Gatekeeper) automatically rejects unsafe workloads at deploy time. Examples: no privileged containers, images only from approved registries, required labels present.
- Supply-chain security — the platform's CI signs each image and produces an SBOM (a bill of materials of what is inside the image), and the cluster only admits signed images. This guarantees every running container actually came from your pipeline, not from somewhere untrusted.
- Least-privilege IAM, Pod Security Admission, and ingress conventions — encoded once in the platform's application abstraction (the kro composition or, on KubeVela, the OAM traits), so every app inherits them by default instead of each team getting them right by hand.
- Git is the default change path, for people and assistants alike: an assistant holding
kubectlmust be told that akubectl patch/edit/scaleon an ArgoCD self-heal resource is reverted on the next sync; it proposes the change in Git, and pauses sync (a scoped, recorded break-glass window) only after the user confirms (see references/gitops-delivery.md).
The platform team owns these controls; app teams get them for free. For the cluster-level depth behind each one, use eks-best-practices (its security.md and security-supply-chain.md references).
Application Modeling — the developer-facing abstraction
The goal of this layer is one declarative file in which a developer requests "an app + a DynamoDB table + an IAM-scoped service account + ingress" and the platform renders everything underneath — never raw Deployments/Services/Ingresses by hand.
Default — Backstage templates + kro. A Backstage software template collects the parameters; a kro ResourceGraphDefinition (authored by the platform team) composes the workload and its dependencies into one custom resource. This keeps the platform on a single composition engine (the same kro used for environment and resource provisioning).
Alternative — OAM / KubeVela. KubeVela implements the Open Application Model: developers describe apps with the OAM Application CRD, composed from platform-authored Components (a runnable unit, e.g. appmod-service, dynamodb-table) and Traits (an operational add-on, e.g. path-based-ingress, component-iam-policy) written in CUE, ordered with dependsOn. It is a richer application model and a fully supported choice. Note its current project velocity before standardizing on it.
Both abstractions, the CUE authoring model, the appmod-service example, and how to choose between them: references/application-model-oam.md.
Progressive Delivery and Promotion
The platform's appmod-service component defaults to an Argo Rollouts canary: 20% → 40% → 60% → 80% → 100%, with quality gates that auto-rollback on failure:
- Functional gate (at 20%) — smoke/correctness test.
- Performance gate (at 80%) — load/latency test (e.g. Artillery).
- Metrics gate — developer-defined Prometheus queries (e.g. avg response time > 3s → fail).
Kargo orchestrates multi-stage promotion: a Warehouse watches ECR for new images; the dev stage auto-promotes; the prod stage requires manual approval and promotes the exact same artifact validated in dev. All promotions are Git commits (auditable, reversible). Strategies (canary/blue-green/A-B), gate config, and the Kargo Warehouse/Stage/Freight model: references/progressive-delivery.md.
Measuring Platform Success
A platform you can't measure is a platform you can't justify. Track the four DORA metrics — deployment frequency, lead time for changes, change failure rate, recovery time — with Apache DevLake ingesting signals from Argo Workflows/Rollouts and GitLab, visualized in Grafana. Measurement is zero-overhead: it's wired in when a team onboards via Backstage.
Pair delivery metrics with cost visibility (showback): attribute spend per team/tenant and surface it in the portal so each team sees what its workloads cost. DORA tells you how fast you ship; showback tells you what it costs to run. Cluster-level cost levers (Spot, Graviton, right-sizing, Karpenter consolidation) live in eks-best-practices (cost-optimization.md).
Framework and dashboards: references/measuring-success.md.
GenAI-Assisted Platform Engineering
GenAI accelerates both tracks: code generation (app developers) and platform generation (platform engineers — Backstage templates, kro/OAM compositions, deployment manifests). The reliable pattern is reference example + target schema + prompt → generated artifact → human review. Always human-in-the-loop; expect hallucinations.
Use Kiro, AWS's spec-driven agentic development tool and the official successor to Amazon Q Developer (Q Developer IDE plugins and paid subscriptions reach end of support on April 30, 2027; new sign-ups closed May 15, 2026). Kiro's spec-driven workflow (structured requirements → design → tasks), steering files for persistent project context, hooks, and custom subagents map directly onto platform-template generation. Spec/steering workflow, prompt patterns, and the migration note from Q Developer → references/genai-platform-engineering.md.
Identity and Multi-Tenancy
Keycloak provides SSO across all tools; for workload AWS access, EKS Pod Identity is the recommended approach for new workloads while IRSA remains a fully supported alternative (and the right choice on Fargate, Windows nodes, or where you already run OIDC federation); per-team namespaces, RBAC, and one-repo-per-component keep tenants isolated. Details: references/identity-and-tenancy.md.
Platform for AI/ML and Data Engineering
The same golden-path model extends to ML and data teams via Backstage templates:
- Model development — JupyterHub (multi-user notebooks, Keycloak SSO).
- Model serving — Ray Serve (Backstage template → Git → ArgoCD → Ray cluster → inference endpoint).
- Data engineering — the Kubeflow Spark Operator (
kubeflow/spark-operator) (Backstage template → Argo Workflows →SparkApplicationCRD). - Key pattern — a DaemonSet pre-pulls large ML images to nodes to kill cold-start latency (on EKS Auto Mode, where nodes are managed and can recycle, pair this with a warm pool / node-lifecycle-aware approach — see reference).
Full ML/data golden paths: references/aiml-data-platform.md.
How to Use the References
This skill uses progressive disclosure — the essentials are above; load a reference only when the task needs that depth:
| Reference | Load when the task is about… |
|---|---|
| idp-architecture.md | IDP concept, reference architecture, hub-and-spoke, GitOps Bridge, value/challenges |
| developer-portal-backstage.md | Backstage portal, software templates/scaffolder, catalog, self-service flow |
| gitops-delivery.md | ArgoCD + Argo Workflows, app-of-apps, cluster registration, CI/CD wiring |
| progressive-delivery.md | Argo Rollouts strategies, quality gates, Kargo promotion |
| infrastructure-abstraction.md | ACK and kro, self-service environment/resource provisioning |
| application-model-oam.md | The developer-facing app model — Backstage+kro (default) vs OAM/KubeVela (alternative), components, traits, CUE |
| golden-paths.md | Golden-path design, guardrails, the developer contract, onboarding |
| identity-and-tenancy.md | Keycloak SSO, Pod Identity/IRSA, multi-tenant isolation |
| measuring-success.md | DORA metrics, Apache DevLake, platform dashboards |
| genai-platform-engineering.md | Kiro (successor to Amazon Q Developer) for templates/manifests, spec/steering workflow, prompt patterns |
| aiml-data-platform.md | JupyterHub, Ray Serve, Kubeflow Spark Operator golden paths |
Sources
- Platform Engineering on EKS workshop (PEEKS) and its companion appmod-blueprints
- Internal Developer Platform · CNCF Platforms White Paper
- Backstage · Argo Project (CD, Workflows, Rollouts) · Kargo
- AWS Controllers for Kubernetes (ACK) · kro · KubeVela / OAM · Kubeflow Spark Operator
- Amazon EKS Auto Mode · EKS Pod Identity · Kiro · Amazon Q Developer end-of-support notice · Apache DevLake (DORA)
이 스킬의 파일
12eks-platform-engineering skill 질문
Claude Code에 eks-platform-engineering skill을 어떻게 설치하나요?
npx degit aws-samples/sample-apex-skills/skills/eks-platform-engineering ~/.claude/skills/eks-platform-engineering를 실행합니다. Claude Code는 ~/.claude/skills(모든 프로젝트) 또는 .claude/skills(한 프로젝트)에서 skills를 로드하며, 설명이 작업과 일치할 때 skill을 사용합니다.
eks-platform-engineering skill을 사용할 수 있는 클라이언트는 무엇인가요?
Agent Skills 형식을 읽는 모든 클라이언트(Claude Code, Codex, Gemini CLI, claude.ai(ZIP으로 업로드))가 사용할 수 있습니다. skill 내 스크립트는 클라이언트가 코드 실행을 허용하는 경우에만 실행됩니다.
EKS 기반 Internal Developer Platform을 위한 권장 도구 스택은 무엇인가요?
권장 스택은 개발자 포털을 위한 Backstage, GitOps 배포를 위한 ArgoCD, 점진적 배포를 위한 Argo Rollouts, 다단계 프로모션을 위한 Kargo, 그리고 AWS 리소스 프로비저닝을 위한 ACK를 포함합니다.
플랫폼 엔지니어링에서 'golden path'란 무엇인가요?
golden path는 정책 기반 코드(policy-as-code) 및 공급망 보안과 같은 가드레일이 내장된, 의도한 소프트웨어가 실행되기까지의 표준화되고 검증된 경로를 의미합니다.