prometheus-query-patterns

What problem does it solve? Writing correct PromQL queries is error-prone: engineers often misuse counters without rate(), pick wrong time windows, or over-aggregate and lose critical dimensions. This Skill provides a reference of proven Prometheus query patterns covering the RED method, SLO tracking, and alerting rules. ## Core Features & Use Cases - Fundamental Operators: Ready-to-use patterns for rate(), histogram_quantile(), and sum() aggregations with label filtering. - RED Method Implementation: Complete Rate, Errors, and Duration queries for HTTP service monitoring, including error-rate percentage calculations. - Alerting & SLO Examples: YAML alert rule templates for high error rates and latency, plus availability and error-budget burn calculations. - Use Case: You need to build a dashboard and alerts for a new API service. Use this Skill to get the correct P95 latency query, error-rate ratio, and alert rule definitions without guessing PromQL syntax. ## Quick Start Ask for the PromQL query to calculate the 95th percentile latency and error rate for your HTTP service, and get a ready-to-use alert rule.