Getting Started with Document Projections
Learn what document projections are, how they generate pages from your model data, and how to create your first flow in the visual flow editor.
Document projections let you generate entire documentation pages directly from your data model. Instead of writing content by hand, you build a visual flow that describes how to traverse your model entities, filter them, loop over collections, and produce structured output blocks such as tables, charts, and rich text. When the site is published, each document projection resolves into one or more content pages automatically.
Why Use Document Projections
How It Works
A document projection is a directed flow graph. Data enters at the top through a source root node, passes through transformation nodes (filters, loops, selectors, decisions), and exits through output nodes (scalar text, tables, charts, responsive rows). During publishing, the flow engine walks the graph, resolves each node against the live model data, and emits native content blocks that the viewer renders.
Key Concepts
Opening the Flow Editor
Creating Your First Flow
A simple flow that lists all entities in your model requires just three nodes:
Node Types at a Glance
| Node Type | Purpose | Category |
|---|---|---|
| Source Root | Entry point, selects the starting data | Source |
| Filter | Narrows data with conditions (AND/OR) | Transformation |
| Loop | Iterates over a collection | Transformation |
| Selector | Picks a specific item | Transformation |
| Decision | Branches the flow based on conditions | Transformation |
| Scalar Output | Emits rich text content | Output |
| Array Output | Emits a table | Output |
| Chart Output | Emits a chart visualisation | Output |
| Responsive Row | Emits a multi-cell row with mixed content | Output |