Skip to main content

culvii workflow run

Start a workflow execution.

Synopsis​

culvii workflow [--env <env>] run <workflow> --trigger <node-name> [--payload <json>] [--payload-file <path>] [--mocks-file <path>] [--json]

Description​

Starts a workflow execution in feedback-mode. This mode allows you to test workflows with simulated inputs and mocked outputs for external services.

  • Triggering: Use --trigger to specify the starting node. If it's a manual trigger, the --payload or --payload-file is injected directly into it.
  • Mocking: The --mocks-file allows you to provide pre-defined outputs for specific nodes, keyed by their step id.
    • Requirement: External connector nodes, core.http, and core.browserAutomation must be mocked in feedback-mode.
    • Unmocked: Core logic nodes (like core.if, core.set) and core.agent cannot be mocked; they always execute their real logic.
  • Mocks File Format: A JSON file containing either a single object (mapping node IDs to outputs), an array of such objects, or a specialized object with a batches key: { "batches": [[...]] }.

Flags​

FlagRequiredDescription
--env-Target environment (dev, sandbox, or prod). Defaults to active context.
--trigger✓Trigger node name.
--payload-Trigger payload as JSON.
--payload-file-Read trigger payload from a JSON file.
--mocks-file-JSON object mapping node IDs to mock outputs.
--json-Print machine-readable JSON.

Examples​

culvii workflow run my-workflow --trigger manual-trigger --payload '{"key": "value"}'

Error behaviour​

ConditionMessageExit
Both payload flags"Use only one of --payload or --payload-file"1
Invalid JSON"<label> is not valid JSON: <error>"1
Invalid mocks file"Mocks file must contain an object keyed by workflow node ID"1
Workflow not found"Workflow not found: <workflow>"1