Dashboards.
Dashboards are saved views composed from cells you've already written. Pin a chart, pin a number, arrange them on a grid, publish when you're ready.
From notebook to dashboard
Every dashboard tile points back at a cell. Re-run the cell, the tile updates. Edit the SQL behind a chart, the chart on the dashboard catches up. No copy-paste, no separate layer of definitions to maintain.
Most teams keep one project-level dashboard.orchid for the headline view and spin up a second one for monthly review or a one-off readout. Both are just files in your project — diff in git, publish on demand.
What lives where
- Building a dashboard — pinning cells, the dashboard file, live links to sources.
- Tile types — Chart, Table, KPI, Markdown.
- Filters — bind a control to a variable, refresh every tile that uses it.
- Sharing — publish modes, custom slugs, auto-refresh.
A quick mental model
A dashboard is a YAML file that points to cell ids. Pinning a cell adds a tile entry to the dashboard; the tile fetches the cell's current output when rendered. That's the whole architecture.
orchid_dashboard: '1.0'
metadata:
title: Monthly revenue review
created: '2026-05-14T09:00:00Z'
tiles:
- id: kpi_total
type: kpi
source:
kind: notebook_block
notebook_id: revenue.orchid
block_id: total_revenue
layout: { x: 0, y: 0, w: 3, h: 2 }
- id: chart_monthly
type: chart
source:
kind: notebook_block
notebook_id: revenue.orchid
block_id: chart_by_month
layout: { x: 3, y: 0, w: 9, h: 5 }Right-click a cell in any notebook and pick Pin to dashboard. The first time, Orchid offers to create one for you. After that, pick which dashboard to add to.
Live vs. snapshot
In Orchid, dashboards are always live against the cells in your project. Re-run the cell, the tile reflects it. Published dashboards are static snapshots by default — you can opt in to scheduled refreshes to keep public viewers in sync.
Next: Building a dashboard, or read about tile types, filters, and sharing.