Automate Application Deployments with CI/CD

Deployment Engineer is an autonomous agent that designs, implements, and optimizes deployment pipelines, containerization, and infrastructure automation. It works through seven stages: requirements analysis (examining application architecture and dependencies, identifying deployment targets, assessing scalability, security, and reliability needs, and documenting current pain points), pipeline design (a CI/CD workflow with build, test, and deploy stages, selecting tools like GitHub Actions, GitLab CI, or Jenkins, choosing a deployment strategy such as blue-green, canary, or rolling updates, and planning rollback and disaster recovery), containerization strategy (optimized multi-stage Dockerfiles, security practices like non-root users and minimal base images, a container registry and image-tagging strategy, health checks and resource limits), Kubernetes configuration (deployment manifests, services, and ingress, ConfigMaps and Secrets management, horizontal pod autoscaling and resource quotas, monitoring, logging, and observability), infrastructure as code (Terraform or Helm charts, environment-specific configs, secrets management and encryption, automated infrastructure testing), implementation and testing (deploying pipeline configs, end-to-end deployment tests, rollback validation, performance testing), and documentation and handoff (runbooks, architecture decisions, monitoring dashboards and alerting, team training).

Use it when you need a complete deployment package - pipeline, containers, Kubernetes resources, infrastructure code, and documentation - designed with security and reliability first. It is not a shortcut around the agent's own guideline of minimizing manual intervention through automation, and it explicitly favors designing for failure (health checks, circuit breakers, monitoring) over assuming deployments will always succeed.

Given an application's architecture and target environments, it outputs a deployment package: pipeline configuration files, container manifests (Dockerfiles and compose files), Kubernetes YAML for all K8s objects, infrastructure code (Terraform or Helm charts with variables), staging and production environment configs, a monitoring setup (Prometheus, Grafana, or equivalent), and setup, deployment, and troubleshooting documentation. It also produces an implementation report covering the architecture overview and data flow, a security assessment, deployment performance metrics, step-by-step rollback procedures, and a maintenance guide for updates and scaling.

### Multi-stage build for optimization
FROM node:18-alpine AS builder
WORKDIR /app
COPY package*.json ./
RUN npm ci --only=production

FROM node:18-alpine AS runtime
RUN addgroup -g 1001 -S nodejs && adduser -S nodejs -u 1001
WORKDIR /app
COPY --from=builder --chown=nodejs:nodejs /app .
USER nodejs
EXPOSE 3000
HEALTHCHECK --interval=30s --timeout=3s --start-period=5s --retries=3 \
  CMD curl -f http://localhost:3000/health || exit 1
CMD ["npm", "start"]

Targets CI/CD platforms such as GitHub Actions, GitLab CI, and Jenkins; container registries; Kubernetes; and infrastructure-as-code tooling (Terraform, Helm), alongside observability stacks like Prometheus and Grafana.

Platform and DevOps engineers who need a reproducible, security-first deployment setup that another team member can operate - built around guidelines for security (scanning, secrets management, least privilege), reliability (health checks, circuit breakers), efficiency (build times, resource usage), scalability (horizontal scaling), observability (logging, metrics, distributed tracing), automation, and documentation.