Viewing a Node's Data Shape in the Workflow Builder

Overview

When writing TypeScript in a node's mapper or condition function, you often need to know exactly what fields an upstream node produces before you can reference them. The View data details panel shows the inferred shape of that data — field names, types, and whether they're required — directly from the chart, without running the workflow.

In this article, we'll be covering how to open the data details panel, how the information is organized, how to search it, and how it differs from the data you see in a workflow execution.

Opening the data details panel

1. Open your workflow in the builder
2. Hover over a node on the chart to reveal its kebab (⋮) menu
3. Click View data details

Reading the data details panel

The panel is split into two sections:

• Node input state — the data available to this node, organized into three groups
• Node output — the data this node produces

Node input state groups upstream data into:

• Latest step — the node or nodes that run immediately before this one
• Previous steps — every other node earlier in the chart
• Context data — state fields outside of steps, such as the trigger's own fields (for example objectId or input)

Each node appears as a card you can expand to see its field tree, with the type of each field. A card with no typed fields shows "The output has no typed fields." A group with nothing to show — for example, a node placed right after the trigger has no previous steps — is hidden rather than shown empty.

If your window is narrow, the Node output section starts hidden. Click Show output to open it, and Hide to close it again.

Some node types don't expose an inferable output shape. In that case, you can set one explicitly by clicking Add output schema in the node's side panel.

Searching the data details panel

Use the search box at the top of the panel to find a field without expanding every card. Search looks at field names and node names, and ignores capitalization and accents. It looks through the whole shape, including fields inside collapsed cards.

1. Open the data details panel.
2. Click the search box, which reads Type to search, and start typing.

As you type:

• Matching text is highlighted, and the cards and groups that contain a match open so you can see it.
• Cards without a match are hidden.
• The panel shows how many results were found in each section, for example "3 results in Node input state" and "1 result in Node output."
• If the Node output section was hidden because your window is narrow, it opens so you can see its matches.

Press Esc to clear the search. Press Esc again to close the panel.

Reading nodes that return a single value

Not every node returns an object with fields. Some return a single value, such as a piece of text or a number, or a list of simple values. For these nodes, the card doesn't show a field tree. It tells you what the node returns instead. For example: "This node returns a string, not a structured object, so there's no data to break down here."

The panel describes these kinds of output:

• A string, number, or boolean
• Null
• Multiple types
• An array of strings, numbers, or booleans, or an array of multiple types

Cards whose output is a list show an array tag next to the node name. A list of simple values doesn't show extra entries such as length, only the values it holds.

Understanding shape vs. actual data

The data details panel shows the shape of your data — field names and types — inferred from node configuration, without running anything. It does not show real values from an execution.

To see the actual data produced by a specific run, use the execution history instead. See Understanding Workflow Executions.

Understanding limitations

• Very deeply nested schemas may not load in full — a warning appears when some fields were left out.
• If the data shape can't be loaded at all, the panel shows an error state.