Data and context

Execution context

Each automation run has an execution context. It starts with signal data and can receive values produced by steps during that run.

Data paths

A path selects a field inside structured context data. Available roots and fields depend on the signal, previous steps and plugin actions used by the automation.

Use a field picker or autocomplete where available. A valid path can still have no value when the expected event field was not supplied.

Step input and output

A step may read literal settings, variables or context paths. A step that produces output can make that result available to later steps in the same execution.

Runtime variables

Runtime variables are temporary values stored under runtime.*. They exist only inside one automation run and are discarded when that run finishes.

Use a runtime variable for intermediate results that later steps in the same sequence need, such as runtime.result.count or runtime.ai.last.text. Concurrent runs have separate runtime state.

Use a project variable under vars.* when a value must remain available to later automation runs.

Value substitution

Some text and JSON fields support value substitution from context paths. Reusable program logic is implemented with macros.

Context storage

Context storage is separate from the in-memory execution context. It provides controlled file access through configured context roots and can be managed with Data Explorer.

Missing values

Conditions and steps may distinguish an absent field, an empty string, zero and a null value. Check the feature reference before assuming they are equivalent.