Schema Overview

Every flowctl YAML file starts with schema: nebula and describes a single flow or subflow.

Root Fields

Field	Type	Required	Default	Notes
schema	string	No	nebula	Schema version. Unknown values produce a clear error.
sys_id	string	No	—	32-char hex. Set by --export --with-uuid. Required for --update.
name	string	Yes	—	Flow display name.
type	string	No	flow	flow or subflow.
status	string	No	draft	draft or published.
description	string	No	—	HTML description (same as Flow Designer).
run_as	string	No	system	system or user.
access	string	No	public	public or package_private.
priority	string	No	—	HIGH or MEDIUM.
active	boolean	No	true	Whether the flow is active.
scope	string	No	—	Application scope sys_id.
trigger	object	Flows	—	Required for type: flow. See Triggers.
inputs	list	No	—	Subflow input parameters. See Variables.
outputs	list	No	—	Subflow output parameters.
variables	list	No	—	Scratch variables.
stages	list	No	—	Service Catalog stages. See Stages.
steps	list	Yes	—	Flow steps. See Step Types.
_metadata	object	No	—	Generated by --export --with-uuid. Don’t edit.

Minimal Flow

schema: nebula
name: Hello World Flow
type: flow
trigger:
  type: record_created
  table: incident
steps:
  - action: log
    inputs:
      message: "A new incident was created: $trigger.current.number"

Full Example

schema: nebula
name: Incident Auto-Assignment
type: flow
status: draft
description: Assigns new P1 incidents to the on-call group.
run_as: system
access: public
priority: HIGH

trigger:
  type: record_created
  table: incident

variables:
  - name: assigned_group
    type: string

steps:
  - action: lookup_record
    id: find_group
    name: Find On-Call Group
    inputs:
      table: sys_user_group
      conditions: "name=On-Call Support"

  - if: "$trigger.current.priority = 1"
    then:
      - action: update_record
        name: Assign to On-Call
        inputs:
          table: incident
          values:
            assignment_group: $find_group.record.sys_id
      - action: send_email
        inputs:
          to: $find_group.record.email
          subject: "P1: $trigger.current.short_description"
          body: "A new P1 incident requires attention."
    else:
      - action: log
        inputs:
          message: "Non-P1 incident, no auto-assignment."

Variables

Inputs, outputs, and variables share the same format:

Field	Type	Required	Notes
name	string	Yes	Must match ^[a-z][a-z0-9_]*$.
label	string	No	Display label.
type	string	No	Data type (see below).
mandatory	boolean	No	Required field.
default	string	No	Default value.
reference	string	No	Reference table name (for type: reference).
max_length	integer	No	Maximum string length.
array	boolean	No	Array variable.
help	string	No	Help text.
hint	string	No	Placeholder hint.

Variable Types

string, integer, boolean, decimal, float, reference, table_name, glide_date_time, glide_date, html, journal, journal_input, url, email, phone_number, script, conditions, document_id, choice, currency, price, sys_class_name, translated_text, ip_addr

Subflow Example

schema: nebula
name: Send Notification
type: subflow
inputs:
  - name: recipient
    type: reference
    reference: sys_user
    mandatory: true
  - name: message
    type: string
    mandatory: true
outputs:
  - name: success
    type: boolean
steps:
  - action: send_email
    inputs:
      to: $input.recipient.email
      subject: Notification
      body: $input.message
  - assign_outputs: "Set result"
    inputs:
      success: "true"

Schema Versioning

The schema field uses celestial codenames. The current version is nebula.

Files without a schema field are treated as nebula. If a file uses a version newer than your tool supports, you’ll see: unknown schema version "xyz"; upgrade flowctl.

When the format changes in a breaking way, a new version will be added with migration guidance.