And Operator

This guide explains how to configure and use the AndOperator node in your automation workflows to join multiple parallel branches into a single merged output.

1. Purpose

• The AndOperator node acts as a join or aggregation point in your workflow.
• It waits for all upstream branches to complete, then merges their outputs into one unified context for downstream tasks.

---

2. Configuration

2.1 Dependencies

• Draw dependency edges from each upstream task into the AndOperator node.
• Use any status branch (e.g. Completed, Success, or custom status names) on each dependency—AndOperator will collect all regardless of branch name.

[TaskA] ──(Completed)──╮
                       ▼
                     [AndOperator] ──(Completed)──▶ [NextTask]
                       ▲
[TaskB] ──(Completed)──╯

2.2 Inputs

• No additional TextInputs are required on the AndOperator node itself—its only job is to merge upstream data.

---

3. Behavior

1. Hold state: As each upstream task finishes, its output variables are stored in the workflow’s global result map.

2. Release: Once every parent task has produced output, the AndOperator node becomes eligible to run.

3. Merge: When executed, AndOperator automatically:

   • Gathers each parent’s output data dictionary.
   • Combines them into one merged dictionary (later values override earlier on key conflicts).
   • Injects merged into its own data and makes them available to downstream placeholders (e.g. , ).

4. Continue: After merging, it sets taskSuccess = true and routes along the configured branch (usually Completed).

---

4. Example JSON Configuration

In this example:

• FetchCustomer and FetchOrders run in parallel.
• After both finish, AndOperator merges their outputs.
• Downstream tasks can reference both and in one place.

---

5. Troubleshooting

• AndOperator never fires: Ensure every upstream task has a dependency route into the AndOperator, and that each has completed successfully.
• Missing data keys: Verify that upstream tasks output the variables you expect (check their data maps) and that there are no naming collisions.
• Early release: AndOperator will not run until all parents appear in the _results map—no special configuration needed beyond dependencies.

---