Diagnose CloudWatch alarms, metrics, and observability gaps across AWS services.
npx skills add https://github.com/LiboMa/agenticops-chat --skill monitoring-liboma
Please help me install this Agent Skill. Skill: monitoring Source: https://github.com/LiboMa/agenticops-chat/tree/main/skills/monitoring Command: npx skills add https://github.com/LiboMa/agenticops-chat --skill monitoring-liboma
High-intent search queries and answers about installing and using this skill.
INSUFFICIENT_DATA usually means the metric is not being published. Verify the metric exists with list-metrics, check that dimensions match exactly (case-sensitive), confirm the CloudWatch agent is running, and ensure the alarm period is not shorter than the metric publishing frequency.
Classify alarms into critical, warning, and info tiers, then increase evaluation periods for non-critical alarms, use composite alarms to combine related signals, and apply anomaly detection bands for variable workloads instead of static thresholds.
Key RDS metrics include CPUUtilization, FreeableMemory, DatabaseConnections, ReadIOPS/WriteIOPS, FreeStorageSpace, ReplicaLag, SwapUsage, and DiskQueueDepth. The Skill provides warning and critical thresholds for each, such as FreeableMemory below 512MB as a warning.
Use put-metric-alarm with a metrics array containing two source metrics (errors and total requests) plus a math expression like (errors/total)*100. Set ReturnData to true only on the expression, then alarm on the computed percentage threshold.
Use anomaly detection for metrics with natural variation or seasonality, such as request counts and latency, where a fixed threshold causes false positives. Static thresholds remain appropriate for hard limits like disk space or status check failures.
itsmostafaAutomate AWS CloudWatch monitoring setup for logs, metrics, alarms, and dashboards.
allthingslinuxRoute alerts and coordinate incidents across PagerDuty, Alertmanager, and Slack.
d0whc3rAnalyze logs and stack traces to identify root causes across distributed systems.
bobmatnycIntegrate Prometheus and Grafana for metrics collection, alerting, and dashboards.
