Write a dashboard specification with AI
Define who needs the dashboard, which decision it serves and what its metrics, filters and acceptance checks must mean in practice.
Write a dashboard specification with AI around a decision someone must make. A branch manager checking tomorrow's rota needs different information from a regional manager comparing four months of staffing costs. The dashboard-specification skill helps settle which audience and question a particular view should serve.
Give the team a document it can test. Define the level at which a number is calculated, how often it is refreshed and what should happen when someone changes a filter. A working example makes disagreements about the metric visible while they are still inexpensive to fix.
What it does
- Identify the primary audience and intended action.
- Separate headline measures from supporting detail.
- Document a metric's source, definition and update schedule.
- Describe what a filter or drill-down should change.
- Agree on a scenario that can be used for acceptance.
Install and use
- Provide the permitted business brief, metric definitions and intended users. Planning the specification needs no live reporting account.
- Ask, “Define a staffing view for one branch. How should tomorrow's unfilled shifts be counted, and how will the manager know when the data last changed?”
- Review the draft with the metric owner and the person making the decision before handing it to the builder.
FAQ
Source
Made by Nimrod Fisher. Original skill: dashboard-specification, published on GitHub under the MIT licence.
View on GitHub