Function Step
A Function step runs one of the platform’s built-in functions — reading and writing key-value buckets, object storage and streams, plus a handful of transformations such as base64, XML/JSON conversion and Handlebars templating.Configuration
The editor has four tabs: General, Function, Response Map and Advanced Settings. The Function tab holds a required Function dropdown listing every function available to your account. Choosing one sets the step’ssubject; the step’s payload is then edited against that function’s own input schema, either through the generated form or with the JSON Editor button.
The built-in functions
Pick these by name from the dropdown. The names below are the invocable ones — a few differ from what you might guess from the category.Key-value storage
Object storage
Streams
Transformations and utilities
Several of these overlap with dedicated steps —
HTTP, Rules, Mapping and Delay all have their own step types, which are easier to configure. Reach for the function form of them only when you need it inside a larger function chain.Worked examples
Store something for later
Read it back
Publish an event
Render a template
Response Map
The Response Map is a JSONPath mapping resolved against the function’s result, and what it produces replaces the result under the step ID:Failure and retries
A function that errors fails the step, and a failed step ends the run. Advanced Settings offers the usual controls —timeout, attempts (total, capped at 5), backoff_ms (a constant wait), ignore_response_codes and flat_map.
Editing a flow changes its draft. Publish to make the change take effect.
Next Steps
Functions Reference
Every function’s parameters and output
Key-Value Functions
Bucket reads and writes
Stream Functions
Publishing and reading events
Nested Flow
The other step with a Response Map