All products

Build a Kibana dashboard with AI

Create a reviewable dashboard definition with Lens panels, filters and controls. Match chart layers to the dataset, then check the saved view in its target space.

Build a Kibana dashboard with AI when an operations or product team needs a set of charts that can be reviewed and maintained together. The kibana-dashboards skill uses the dashboard and visualisation APIs to create compact JSON definitions. Dataset choice comes first, because a data view and an ES|QL query call for different panel configuration.

For a sales review, a daily order trend and a category breakdown answer different questions. Use the appropriate chart and fields for each one, then test the time filter across both. Keeping the definition in version control helps reviewers spot a changed aggregation before a dashboard is replaced.

What it does

  • Build a dashboard from a known data view.
  • Create a Lens panel or reuse a saved visualisation.
  • Group trends and comparisons into useful sections.
  • Check dashboard controls against the intended panels.
  • Review the changes in a compact JSON definition.

Install and use

  1. Install the skill and connect official elastic CLI 0.4 or later to Kibana 9.4 or later with matching Elasticsearch. Use an authorised account with the relevant data and saved-object permissions.
  2. Ask, “Build a test-space sales dashboard with orders by day and revenue by category. Verify the dataset and panel fields first.”
  3. Open the saved dashboard, compare the figures with source data and confirm the target before overwriting an existing view.

FAQ

Source

Made by Elastic. Original skill: kibana-dashboards, published on GitHub under the Apache-2.0 licence.

View on GitHub