DevOps Strategy
Generate comprehensive DevOps strategy with CI/CD, Infrastructure as Code, and platform engineering.Command
Description
Defines how software will be built, tested, deployed, and managed throughout its lifecycle. Establishes engineering practices, tooling, and automation for rapid, reliable delivery.Arguments
- Project ID or platform: Project identifier or platform description (e.g., ‘001’, ‘GitHub Actions on AWS’)
When to Use
- After requirements and architecture diagrams
- Before implementation begins
- For platform engineering initiatives
- CI/CD pipeline design
- Infrastructure automation planning
Required Context
- Requirements (ARC--REQ-.md) - MANDATORY
- Architecture Principles (ARC-000-PRIN-*.md) - MANDATORY
- Architecture Diagrams (DIAG) - Recommended
- Technology Research (RSCH/AWSR/AZUR) - Recommended
DevOps Maturity Levels
| Level | Characteristics | Deployment Frequency |
|---|---|---|
| Level 1 | Manual builds, scripted deploys | Monthly |
| Level 2 | CI automation, manual deploys | Weekly |
| Level 3 | CI/CD automation, staging gates | Daily |
| Level 4 | Continuous deployment, feature flags | Multiple/day |
| Level 5 | GitOps, self-healing, platform | On-demand |
Document Sections
Section 1: DevOps Overview
- Strategic objectives
- Maturity level (current and target)
- Team structure (platform team, dev teams)
- Key stakeholders
Section 2: Source Control Strategy
Repository Structure:- Monorepo vs multi-repo decision
- Repository naming conventions
- Access control and permissions
- GitFlow / Trunk-based / GitHub Flow
- Flowchart diagram (NOT gitGraph - use flowchart)
- Protected branches and merge rules
- Required approvers
- Review checklist
- Automated checks (linting, tests)
- Format:
type(scope): message - Types: feat, fix, docs, refactor, test, chore
Section 3: CI Pipeline Design
Pipeline Architecture:- Stages: Build → Test → Analyze → Package
- Parallel job execution
- Artifact versioning
- Language-specific build tools
- Dependency management
- Build caching
- Unit tests (80% coverage target)
- Integration tests
- E2E tests
- Contract tests
- Linting (ESLint, Pylint, etc.)
- Formatting (Prettier, Black)
- Code coverage threshold
- Security scanning
- SAST (Static Application Security Testing)
- Dependency scanning (Snyk, Dependabot)
- License compliance
Section 4: CD Pipeline Design
Deployment Pipeline:- Stages: Dev → Staging → Production
- Environment promotion gates
- Approval requirements
- Blue-Green deployment
- Canary deployment
- Rolling deployment
- Feature flags integration
- Manual approval for production
- Automated smoke tests
- Compliance checks
- Automated rollback triggers
- Rollback execution steps
- Rollback verification
Section 5: Infrastructure as Code
IaC Tool Selection:- Terraform / Pulumi / CloudFormation
- Module/component structure
- Reusable modules
- Remote state backend (S3, Terraform Cloud)
- State locking
- State file encryption
- Vault / AWS Secrets Manager / Azure Key Vault
- Secret rotation policies
- Application secret injection
- Scheduled drift detection
- Remediation workflow
Section 6: Container Strategy
Container Runtime:- Docker / containerd
- Base image strategy
- Multi-stage builds
- ECR / ACR / GCR / Docker Hub
- Image tagging strategy
- Image retention policies
- Vulnerability scanning (Trivy, Clair)
- Image signing (Cosign)
- Minimal base images (Alpine, Distroless)
Section 7: Orchestration (if applicable)
Kubernetes/ECS/AKS/GKE:- Cluster architecture
- Namespace strategy
- Resource limits and quotas
- Service mesh (Istio, Linkerd)
- Ingress/networking
- ArgoCD / Flux
- Git as source of truth
- Automated sync
Section 8: Environment Management
Environment Types:- Dev (feature branches)
- Staging (pre-production)
- Production (live)
- Configuration as code
- Infrastructure templates
- Data management across environments
- PR preview environments
- Auto-cleanup policies
Section 9: Developer Experience
Local Development:- Dev containers / devcontainers
- Docker Compose for local stack
- Local environment setup automation
- Fast feedback loops
- Hot reload / live reload
- Local testing capabilities
- Platform API or portal
- Service templates
- Automated provisioning
Section 10: Observability Integration
Logging:- Centralized logging (ELK, Splunk, CloudWatch)
- Structured logging format
- Log retention policies
- Prometheus / CloudWatch / Datadog
- Custom metrics instrumentation
- Dashboard provisioning as code
- Jaeger / X-Ray / Zipkin
- Distributed tracing
- Span correlation
- Alert rules as code
- Escalation policies
- On-call rotation
Section 11: DevSecOps
Shift-Left Security:- Security scanning in CI
- Pre-commit hooks
- Security training
- SAST: SonarQube, Semgrep
- DAST: OWASP ZAP
- SCA: Snyk, Dependabot
- Container: Trivy, Clair
- Infrastructure: Checkov, tfsec
- Policy as code (OPA, Sentinel)
- Compliance scanning
- Audit trails
Section 12: Release Management
Versioning:- Semantic Versioning (SemVer)
- Automated version bumping
- Git tags
- Automated changelog generation
- Conventional Commits
- Release branches
- Release notes
- Hotfix process
Section 13: Platform Engineering (if applicable)
Internal Developer Platform:- Self-service portal
- Golden paths / templates
- Platform APIs
- Service catalog
Section 14: UK Government Compliance (if applicable)
- Cloud First (TCoP Point 5) implementation
- Open standards (TCoP Point 4)
- Secure by Design integration
- Digital Marketplace compatibility
Section 15: Metrics & Improvement
DORA Metrics:- Deployment frequency
- Lead time for changes
- Mean time to recovery (MTTR)
- Change failure rate
- Build success rate
- Test coverage
- Code review time
- Retrospectives
- Metrics review cadence
- Automation opportunities
Section 16: Traceability
- Requirements to DevOps element mapping
- NFR-P → Build speed targets
- NFR-S → Auto-scaling
- NFR-SEC → Security scanning
Output
Creates:projects/{project}/ARC-{PROJECT_ID}-DEVOPS-v1.0.md
Example
- DevOps Maturity: Level 3 (target: Level 4)
- Cloud Provider: AWS
- Deployment: EKS (Kubernetes)
- CI: GitHub Actions
- IaC: Terraform
- Container Registry: ECR
- Orchestration: EKS
- SAST: SonarQube
- Target: Daily deployments
Related Commands
arckit finops- Cloud cost management strategyarckit servicenow- Operational ITSM integrationarckit diagram- Deployment topology
Next Steps
After DevOps strategy:- Set up source control repository
- Implement CI pipeline
- Provision infrastructure with IaC
- Configure CD pipeline
- Set up secret management
- Establish DORA metrics baseline
- Train team on platform
Key Principles
- Automation First - Automate everything possible
- Security Shift-Left - Security in CI, not just production
- Infrastructure as Code - No manual changes
- Developer Experience - Fast feedback, self-service
- Observability by Default - Logging, metrics, tracing from day one