Purpose
The block ends a selected subflow processing path and defines the table returned to the parent’s “Call flow” block. One child can declare several outputs, such as cleaned rows and totals. A call chooses the output it needs.
Input and contract
Connect the processing result to the input. Set fields to the fields to return, or leave the list empty for all fields. This list does not change cell values; perform calculations and renaming earlier. Contract field names must be unique.
Designing the output
Give the output a meaningful name. For a stable contract, list fields in the required order. Up to 500 fields are supported; names are at most 256 characters and cannot use the reserved __bridge_ prefix. Create separately named outputs for different detail levels.
Setup
- Add the block after the intended transformation.
- Set its name and field list.
- Ensure its dependency path uses permitted subflow blocks.
- Publish the child flow.
- Select this output in the parent call and inspect the returned table.
After schema changes
Renaming or removing fields changes the contract for subsequent steps. Check parent filters, formulas, joins and destinations after a change. Deleting the output itself requires reselection in the call. This block does not publish an HTTP address or send a table to an external system; configure separate destinations or an API publication for those actions.
Two-result example
After cleaning, create “Rows” with Product, Period and Amount. A parallel Group branch produces “Totals” with Period and Total. A detailed load chooses Rows; a management-report process chooses Totals. Both use the same published cleaning logic.