Generators
Reusable generators for JSON objects, strings, or random byte payloads that can be referenced across Datagen tasks.
Datagen
CertifiedTasks that produce synthetic data payloads for triggering and exercising Kestra workflows.
Use Datagen to simulate events without external sources by pairing core tasks or triggers with JSON, string, or byte generators. Configure a generator (value template or size) and optionally batch and store results so downstream tasks can read generated files from Kestra storage.
Datagen
Certified
Tasks that produce synthetic data payloads for triggering and exercising Kestra workflows.
Use Datagen to simulate events without external sources by pairing core tasks or triggers with JSON, string, or byte generators. Configure a generator (value template or size) and optionally batch and store results so downstream tasks can read generated files from Kestra storage.
How to use the Datagen plugin
Generate synthetic test data — JSON objects, strings, or random bytes — to drive and exercise Kestra flows without external data sources.
Tasks
The generating tasks (core.Generate, core.Trigger, core.RealtimeTrigger) require a generator — choose from JsonObjectGenerator (set value as a map of field names to Datafaker expressions, e.g. #{name.firstName}), StringValueGenerator (set value as a string with Datafaker expressions), or RandomBytesGenerator (set size as the byte count). Optionally set locale on JSON and string generators to control Faker locale.
core.Generate produces data once — set batchSize to generate multiple records (default 1). Set store: true to write results to internal storage instead of returning them inline.
core.Trigger polls on a schedule (default every 1 second) and starts one execution per batch. Set batchSize and store the same way as core.Generate.
core.RealtimeTrigger generates records continuously — start one execution per record as it is produced. Control throughput with throughput (records per second, default 1) and cap total records with maxRecords (default unlimited).
core.Pseudonymize does not generate data — instead it replaces PII in an existing file. Set from (the URI of a CSV, JSON, or ION file in internal storage) and fields (a map of field paths to Datafaker expressions). For JSON and ION, use dot-notation paths to reach nested values (e.g. user.profile.email); for CSV, use column header names. Any field not listed is passed through unchanged. Set contentType (CSV, JSON, or ION) explicitly, or let it be auto-detected from the file extension.
