Building Dashboards

Dashboard building with Agent

The Agent can build private Stackless dashboards from existing Semantic Models, chart cards, table artifacts, or a natural-language dashboard request.

The best Agent-built dashboards are model-backed. If the right Transformation Model or Semantic Model does not exist, the Agent should identify the missing layer instead of forcing a dashboard on raw data.

[picture 1. Agent creating a dashboard draft in the artifact workspace]

Start with the business purpose

Begin with the decision or workflow the dashboard should support:

Build a weekly executive dashboard for revenue, orders, AOV, active customers, and channel performance. It should help leadership see whether growth is healthy and which channel needs attention.

Then specify:

• Audience.
• Time period.
• Metrics.
• Breakdowns.
• Filters.
• Tabs or sections.
• Whether it should be private or published.

Ask the Agent to inspect existing models first

Before building, ask:

Which Semantic Models and Transformation Models can support this dashboard?

This helps avoid duplicate modeling work and keeps dashboards tied to governed metrics.

A good dashboard request

Create a private dashboard called Marketing Performance.

Use existing governed models where possible.

Include:
- KPI cards for spend, net revenue, ROAS, orders, conversion rate, and new customers.
- A weekly trend of spend, revenue, and ROAS.
- A channel table with spend, revenue, ROAS, orders, conversion rate, AOV, CPA, clicks, CPC, impressions, and CPM.
- Filters for date range and channel.

Before publishing, show me the model dependencies and any missing data.

Use chart cards as stepping stones

If you are unsure about a visualization, ask for a chart card first:

Create a chart card of weekly net revenue by channel for the last 90 days.

After reviewing it:

Copy this chart card to a new dashboard and add a KPI card for total net revenue.

Chart cards are useful because they let you validate one widget before committing to a full dashboard layout.

Draft, review, publish

Agent-built dashboards should follow this lifecycle:

1. Create a private draft.
2. Preview widgets and filters.
3. Verify metric definitions and lineage.
4. Check model dependencies.
5. Review access and audience.
6. Publish when the dashboard is ready for recurring use.

Publishing is explicit. Ask the Agent to prepare the publish, then review the dependencies and confirmation details.

Editing published dashboards

When editing a published dashboard, Stackless creates an edit fork. Work in the fork, review the changes, and publish the replacement when ready.

Good prompt:

Fork the published Executive Overview dashboard for editing, add a Customer Retention tab, and show me the model dependencies before publishing.

Filter guidance

Be explicit about filters:

• Date range.
• Location.
• Channel.
• Campaign.
• Product.
• Customer segment.
• Country or region.

If filters should apply to every widget, say so. If a filter should apply only to one tab or group, say that too.

Common dashboard pitfalls

Building on raw tables

Raw tables are often not stable business surfaces. Ask the Agent to use governed Semantic Models or create the missing Transformation Model first.

Missing grain

If you ask for "customer-level revenue" but the dashboard uses order-level rows, totals may duplicate. Ask the Agent to state the grain.

Unclear metric definitions

"Revenue" may mean gross revenue, net revenue, booked revenue, collected revenue, or another workspace-specific definition. Ask the Agent to use the existing metric definition or explain the formula.

Publishing too early

Private drafts are for iteration. Publish only after dependencies, filters, metrics, and access are reviewed.

Useful dashboard prompts

• "Build a dashboard from these chart cards."
• "Add a date range filter that applies to every widget."
• "Explain the metric definitions used on this dashboard."
• "Trace all upstream models for this dashboard."
• "Which widgets may double-count because of joins?"
• "Create a scheduled weekly export for this dashboard if my role allows it."