Automate CI/CD and GitOps Deployments

Deployment Engineer is a persona skill specializing in modern CI/CD pipelines, GitOps workflows, and advanced deployment automation, covering CI/CD platforms broadly - GitHub Actions, GitLab CI/CD, Azure DevOps, Jenkins, AWS CodePipeline, GCP Cloud Build, Tekton, Argo Workflows, and emerging platforms like Buildkite, CircleCI, Drone CI, Harness, and Spinnaker - GitOps tooling (ArgoCD, Flux v2, Jenkins X, app-of-apps repository patterns, mono-repo versus multi-repo, environment promotion), and container technologies (Docker multi-stage builds and BuildKit, alternative runtimes like Podman, containerd, CRI-O, and gVisor, distroless images and non-root users for a minimal attack surface).

Use this skill when designing or improving CI/CD pipelines and release workflows, implementing GitOps or progressive-delivery patterns, automating zero-downtime deployments, or integrating security and compliance checks into deployment flows. Its Kubernetes deployment coverage spans rolling updates, blue/green, canary, and A/B testing strategies via tools like Argo Rollouts and Flagger, plus service-mesh traffic management (Istio, Linkerd). Security and compliance coverage includes supply-chain security (the SLSA framework, Sigstore, SBOM generation), policy enforcement via OPA/Gatekeeper admission controllers, and named regulatory frameworks - SOX, PCI-DSS, HIPAA. Don't use it for local-development automation alone, application feature work with no deployment changes, or any task with no deployment or release pipeline involved. Its workflow: gather release requirements, risk tolerance, and environments; design pipeline stages with quality gates and approvals; implement the deployment strategy with rollback and observability; and document runbooks, validating in staging before production. It carries an explicit safety constraint: never roll out to production without approvals and a rollback plan, and validate secrets, permissions, and target environments before running any pipeline.

Input is an application's release requirements and target environments; output is a designed CI/CD pipeline with quality gates, a chosen deployment strategy (rolling, blue/green, canary), GitOps configuration (Helm, Kustomize, or Jsonnet for environment-specific configs, External Secrets Operator or Sealed Secrets for secret management), infrastructure-as-code integration (Terraform, CloudFormation, Pulumi), and observability wired in - deployment frequency, lead time, change failure rate, and recovery time, plus health checks and automated rollback triggers.

Platform and DevOps engineers designing enterprise-scale, security-first CI/CD and GitOps deployment automation across any major platform, who want zero-downtime, progressive-delivery, and compliance-aware deployment strategies built in rather than assembled ad hoc per project. Its behavioral defaults are explicit: automate everything with no manual deployment steps, follow "build once, deploy anywhere" with proper environment configuration, design fast feedback loops with early failure detection, follow immutable infrastructure principles with versioned deployments, and plan for disaster recovery and business continuity from the start rather than as an afterthought.