Canvas

A canvas is Spark's dashboard surface: the place where raw data becomes a picture a team can read at a glance. Where a data object is a table of records, a canvas is what you build on top of it to actually see what's going on.

Widgets on a canvas

A canvas is made of widgets, and each widget is one element of the picture. A widget can be a table, a chart, a KPI tile, or embedded tasks and files, among others. You arrange them on the canvas to tell a story: the headline number up top, the trend beside it, the detail below.

Built on live data

The important part is that widgets read from your data objects and views, not from a static snapshot. When the underlying records change, the canvas reflects it. Point a chart at a view and it always shows the intended slice of data, kept current automatically.

That live connection is what makes a canvas a working instrument rather than a report you have to rebuild each week.

Building versus reading

A canvas has two modes. Most of the time you're reading it. When you want to change it, you enter an edit mode to add and arrange widgets, then save. Keeping the two separate means a shared canvas stays stable for everyone reading it while one person reshapes it.

How it relates to the rest

A canvas is a lens, like a view, but a richer, composed one. It sits on top of your data, often lives inside a Space for a use case, and gives both your team and the Agent a shared picture to reason about.

The Canvas section covers how to build one and work with widgets.