Variables & Data Flow

A workflow is a series of steps that pass data to each other. Variables are how that data travels: each step can read values produced by earlier steps and add its own for later steps to use. Understanding variables is the key to building workflows that react to real data instead of fixed values.

Where values come from

Every value in a workflow falls into one of four groups:

• Node - values produced by a step when it runs, such as the phone number from an outbound call or the outcome of a Check node. These become available only after that step has run.
• Config - fixed settings you define on the workflow itself, like a timeout or a default sender name.
• Secret - sensitive settings such as API keys, kept separate so they can be handled carefully. See Secrets.
• System - details the workflow fills in automatically at the start of a run, like the organization it belongs to and the time it started.

Referencing a value

You point at a value by naming its group, the step it came from, and the value itself:

node.outbound_call.caller_phone_number
config.timeout
secret.crm_api_key

In practice you rarely type these out. The builder offers a picker so you can choose a value from a list rather than remembering its exact name.

Values become available in order

A step can only read values that already exist when it runs. Because the connections in a workflow define the order of steps, the builder knows which values are ready at each point and will only offer those.

In this example, steps B and C can both read the value that A produced, but A cannot read anything from C, because C has not run yet:

  A  (produces x)
  |
  B  (can read A's x)
  |
  C  (can read A's x)

Related