This skill designs a product metrics dashboard in layers: a North Star Metric, three to five input metrics, health guardrails and business metrics. Each metric gets an exact definition, a data source, a visualization, a target and an alert threshold.
It draws on the Lean Analytics criteria for a good metric (understandable, comparative, a ratio or rate, behavior-changing) and separates vanity from actionable and leading from lagging metrics. The output is a dashboard specification in markdown.
What You Get
- A layered metric set: North Star, input metrics, health metrics and business metrics
- A table with definition, data source, chart type, target and alert threshold per metric
- A dashboard layout sketch
- A review cadence and alert plan with owners and response times
Common Use Cases
- Setting up product analytics for a new launch
- Auditing an existing dashboard for vanity metrics
- Turning OKRs into a set of tracked metrics with alerts
What Makes It Unique
It embeds a short metrics theory (four criteria for a good metric, eight metric types, five action steps) so the dashboard is checked against whether each number would change behavior. It also fixes a review cadence and alert ownership, which many dashboard templates leave out.
Under The Hood
The skill is a prompt only, with no scripts. It tells the agent to read any dashboards, analytics exports, OKRs or strategy files the user supplies, and to save the specification as a markdown document. It names analytics and dashboard tools as suggestions but does not call any of them or any web service.