kibana-lens

Kibana Lens

Use when the user wants to create or customize visualizations (charts, tables, gauges) in Kibana using Lens.

1. Prompt

• Clarify: data source (data view or ES|QL query), metric or breakdown fields, and chart type (bar, line, pie, table, gauge, etc.).
• Use kibana_list_data_views when KIBANA_URL is set to suggest a data view.

2. Provision

• Ensure the data view exists and has the needed fields; use get_mappings or index APIs if defining new indices.
• Point to Lens via Dashboard (Add panel → Lens) or Discover (Visualize → Create Lens visualization).
• For programmatic creation, use kibana_create_dashboard (see kibana-dashboard skill) — it wraps Lens panels in the as-code API format.

3. Integrate

• Guide: add a Lens panel, choose visualization type, drag fields to dimensions and metrics, adjust formatting.
• For dashboards: use kibana_list_dashboards to list existing dashboards; suggest adding the Lens chart to one.

Display Quality Tips

• Column aliases are labels: In ES|QL-powered Lens panels, the column alias from STATS becomes the axis label, subtitle, and table header. Use `Avg pH` not avg_val.
• ROUND() values: KPI metrics and gauges look cleaner with ROUND(AVG(field), 1) than raw decimals.
• Chart type choice: Line for precision/trend metrics (chemical levels), area for flow/continuous data (water systems), bar_stacked for volume/count data.
• Breakdown legends: Rename dimension columns (site = \`attributes.site.name\`) so legends show "Hospital" not "attributes.site.name".

4. Validate

• Confirm the visualization renders and updates with time range and filters.
• Check that axis labels and legends are human-readable (not raw field names).
• Recommend saving the Lens visualization and adding it to a dashboard (see kibana-dashboard skill).