Skip to Content
Raydo

Data and flow control nodes

Transform inputs, set edge conditions, and inspect why a branch was skipped.

Use Data transform for a declared JSON transform. The default data.transform specification is a passthrough. Select an operation such as pick, rename or filter, then check its required source shape and options. For example, a transform with outputKey set to cleanRows exposes its result at $.nodes.node_clean.output.result.cleanRows. Array filters require an array and a declared predicate.

A Condition branch node helps organize a split on the canvas. Put the test on the outgoing connection, for example $.run.input.includeDraft == true. Add a separate edge for the false case if it has work to do. When every incoming condition evaluates to false, the runtime skips the receiving node. A node with multiple active parents uses the join policy in its configuration.

Flow operation defaults to the capability ID workflow.branch in the editor. Use the readiness shown for the specific route and use conditional edges for ordinary branching. An editor node or imported label alone cannot establish loop, batching or retry support.

If a reference breaks, inspect the upstream output in the run. Run inputs use paths such as $.run.input.includeDraft. Node outputs use paths such as $.nodes.node_clean.output.result.cleanRows. The referenced node must be an upstream dependency.

On this page

Data and flow control nodes