Parameters and Return Values
Steps don’t work in isolation - they read some values, do something, and produce new ones. Parameters are a step’s inputs; return values are its outputs. Wiring them together is how data flows from one step to the next.
Parameters - a step’s inputs
A parameter is a value a step needs to do its job. When you configure a step, you connect each parameter to a piece of process data (a symbol) or a fixed value.
Map only what the step actually needs - extra inputs just add noise.
Return values - a step’s outputs
A return value is something the step produces. You map each return value to a symbol so later steps can use it. Anything you don’t map back isn’t kept.
Keep return-value names and types stable when downstream steps depend on them - a later step reads the symbol you wrote, so renaming or retyping it can break the flow.
Processes have an interface too
A whole process exposes its own Parameters and Return Values - the inputs it’s started with and the results it hands back. This matters when one process runs another as a sub-process: the caller passes values into the child’s parameters and reads its return values back, exactly like calling a single step.
Outcomes are not return values
Don’t confuse return values (the data a step produces) with its outcome (which path it takes - success, failure). Outcomes route the flow; return values carry the data. See Branching & Merging and Error Handling.
Related
- Symbols, Variables & Constants - what the values you map actually are.
- Run Script - maps extra inputs and outputs explicitly.