This is a properly opinionated take on PromQL generation. Instead of just spitting out queries, it walks you through an interactive planning workflow: understand what you're monitoring, identify your metrics and labels, set parameters like time ranges and thresholds, then present a plain-English plan before writing any code. It's designed to catch the common mistakes (like using rate() on gauges or averaging quantiles) and ensure the query actually matches what you need. If you're building Grafana dashboards or Prometheus alerts and want to avoid the typical trial-and-error cycle of PromQL, this handles the thinking part before the syntax part. The workflow might feel slow if you just want a quick query, but that's the point.
npx -y skills add akin-ozer/cc-devops-skills --skill promql-generator --agent claude-codeInstalls into .claude/skills of the current project.
Select a file.
sickn33/antigravity-awesome-skills
kubesphere/kubesphere
supercent-io/skills-template