aws-ecs-monitor
by briancolinger
DevOps & CloudBrief
AWS ECS production health monitoring with CloudWatch.
Description
What This Skill Does
Monitors AWS ECS service health by running HTTP probes, checking ECS task counts, ALB target group status, and SSL certificate expiry. Pulls CloudWatch logs and categorizes errors by type — panics, OOM, timeouts, 5xx responses, and container restarts. When health checks fail, it can automatically trigger log analysis to identify the cause.
Combines HTTP health probes, ECS status, ALB target checks, SSL expiry, and CloudWatch log analysis in one workflow instead of navigating multiple AWS Console pages.
When to Use It
- Checking ECS service status after a failed deployment
- Finding OOM or panic errors in container logs
- Verifying SSL certificate expiry before renewal deadlines
- Detecting repeated container restarts in production
- Getting a cross-service error summary after an incident
Example Workflow
User asks: Why is my API returning 5xx errors?
- 1
Runs ecs-health.sh against the cluster to check ECS service status and ALB target group health
- 2
Detects unhealthy targets and triggers auto-diagnose via cloudwatch-logs.sh
- 3
Pulls CloudWatch logs for the failing service over the past 60 minutes
- 4
Categorizes errors into types (http_5xx, timeout, connection_error, etc.) and filters out ALB health check noise
- 5
Reports error counts per category and flags any container restarts
Error breakdown showing 47 http_5xx responses, 3 connection_error entries, and 2 container restarts in the last hour
Requirements
- AWS account with an ECS cluster
- AWS CLI configured with credentials (access key or IAM role)
- IAM permissions: ecs:ListServices, ecs:DescribeServices, elasticloadbalancing:DescribeTargetGroups, elasticloadbalancing:DescribeTargetHealth, logs:FilterLogEvents, logs:DescribeLogGroups
Installation
clawhub install briancolinger/aws-ecs-monitorRun the CLI command, or paste the prompt into Claude Code, Cursor, Codex or whichever coding agent you use.