All setups
DataSetupAI-generated
Dashboard Spec Writer
Translate a business goal into a complete dashboard specification
What it is
Translate a business goal into a complete dashboard specification
Produce a structured dashboard spec from a goal and audience — chart panels, metric definitions, grains, filters, alert thresholds, and data freshness requirements. Skips straight to implementation code only after the spec is agreed on.
What you'll get
Custom instructionsCompiled from your answers, editable before you export.
Dashboard spec reference cardGives Claude a structured template for producing dashboard specs — panel list, metric definitions, filter requirements, freshness rules — so outputs are consistent and actionable for engineers who build the dashboard.
Export bundle for Claude ProjectsCopy-paste blocks with a guided walkthrough.
Try it on
Weekly feature adoption dashboard for product teamscenario · built in
Claude should produce a structured spec with at least four panels: a primary WAU per feature chart, a week-over-week change table, a per-feature trend sparkline, and a new-vs-returning breakdown. It must define "weekly active feature user" explicitly before placing it on any panel. It should include a feature filter, a date range picker defaulting to the last 12 weeks, and call out whether the events table must have a feature_name or feature_id column.
Executive revenue dashboard specscenario · built in
Claude should produce a spec with clear metric definitions for each of the four metrics (MRR, new ARR, churn rate, NRR) before describing any panel. The spec should note that the Monday morning cadence implies a weekly data refresh requirement. Panels should include month-over-month and year-over-year comparisons. It should flag which metrics require subscription-level data and ask whether trial revenue is included or excluded.
Compiled preview
You are a dashboard spec writer. Produce a complete, structured spec — not implementation code.
Dashboard goal: {{dashboardGoal}}
Key metric: {{keyMetric}}
Primary audience: {{primaryAudience}}
Refresh cadence: {{refreshCadence}}
Rules:
1. Define every metric in the spec in plain English before placing it on a panel.
2. State the grain (one row = one ...) and default date range for each chart panel.
3. List required filters, a data freshness requirement, and alert thresholds for the key metric.
4. Never assume data exists that has not been described; flag missing data needs as open questions at the end of the spec.