What problem does it solve? Deploying containers on AWS involves many failure-prone details: invalid Fargate CPU/memory combinations, wrong networking modes, confused IAM roles, slow deployments from default deregistration delays, and silent log loss. This Skill encodes the correct configurations, gotchas, and troubleshooting steps so container deployments on ECS, Fargate, and ECR succeed on the first attempt. ## Core Features & Use Cases - Task and Service Authoring: Create valid Fargate task definitions with correct CPU/memory pairs, awsvpc networking, secrets injection, and separated execution/task IAM roles. - Deployment and Scaling: Configure ALB integration, health check grace periods, deployment circuit breakers, blue/green deployments, and SQS-based worker scaling. - Debugging and Operations: Use ECS Exec for interactive container debugging, resolve CannotPullContainerError, OOM exit code 137, 502 Bad Gateway, and TargetNotConnectedException. - ECR Management: Set up repositories, lifecycle policies, image scanning, and cross-account image pulls. - Use Case: A developer says "deploy my container app behind a load balancer." The Skill produces a CDK ApplicationLoadBalancedFargateService with circuit breaker rollback, 30s deregistration delay, and auto-scaling, avoiding the common misconfigurations that cause failed deployments. ## Quick Start Ask the assistant to deploy your container image as a Fargate service behind an ALB with auto-scaling and rollback protection.