Automations - Administrator Guide - 6.10 - Cortex XSOAR - Cortex - Security Operations

Cortex XSOAR Administrator Guide

Cortex XSOAR
Creation date
Last date published
Administrator Guide

Create and edit an automation in Cortex XSOAR, including detach and attach, automation settings, etc.

Automations perform specific actions and are comprised of commands, which are used in playbook tasks and when running commands in the War Room.

In the Automation page, you can view, edit, and create automations in JavaScript, Python, or PowerShell. When creating an automation, you can access all Cortex XSOAR APIs, including access to incidents, investigations, share data to the War Room, etc. Automations can receive and access arguments, and can be password protected.

You can use the Script Helper when creating an automation, which provides a list of available commands and scripts, ordered alphabetically. For more information about the automations used in Cortex XSOAR, see the List of Automations (Scripts).

Automation Permissions

Permissions for automations in Cortex XSOAR are determined by the Run as and Role fields. Run as determines the permissions with which the automation runs. Role determines who the automation can be seen and executed by.

By default, most automations in Cortex XSOAR are run as the limited user. However, some automations require higher permissions and are delivered, out-of-the-box with Run as set to DbotRole, which means the automation executes with elevated permissions. When an automation runs with elevated permissions, its results can be viewed even by users with lower permissions. To mitigate this, you should assign a role to the automation that is aligned with the users to whom you want to expose the information the automation can extract.


As content packs may use a number of these automations, and the automations could have some dependency on each another, it is important that you assign the Run as and Role parameters consistently to ensure that all of the automations can run properly.

For example, you have an automation that searches for all incidents in the system. The following scenarios show how the system behaves given the various configurations.

Scenario 1

Set the automation with Run as, defined as DBotRole and do not set the Role parameter.

Anyone can run the automation, whether they are an analyst or administrator, and they have access to all incidents that are returned.


Scenario 2

Set the Run as parameter to DBotRole, and set the Role parameter to Analyst. Any user with at least analyst permissions will have access to execute the automation. All other users are not able to implement the automation in playbook tasks. They cannot run the automation manually, neither from the command line nor in a playbook.

All users can see the results of the execution in the War Room.


Scenario 3

Set the Run as parameter to Analyst and do not set a role. Everyone can run the automation, but only incidents that are viewable by the analyst role are returned.


Scenario 4

You set the Run as as parameter to Analyst and define the Role parameter as Analyst. Only users with at least the analyst role are able to execute the automation. Only incidents that are viewable by the analyst role are returned.


For example, you implemented the following roles on your incidents:

  • Incident 1 viewable to Administrators

  • Incident 2 without any role assigned

  • Incident 3 viewable by Analysts

  • Incident 4 viewable by nested Analyst (custom role)

As a nested analyst, you run the automation to search for the incidents in your system.

  • In scenario 1, you will see incidents 1 through 4.

  • In scenario 2, you will not be able to execute the automation, but if a playbook runs the automation, you will see incidents 1 through 4.

  • In scenario 3, you will be able to see incidents 2 through 4.

  • In scenario 4, you will not be able to execute the automation, however, if a playbook runs the automation, you will see incidents 2 and 4.


When you change permissions for an automation, those permissions do not affect playbooks that are already using the automation. The playbooks continue using the previous permissions even when the playbook is triggered manually.

For more information, see Role-based Permission Levels.

Detach and Attach Automations

When installing an automation from a content pack, by default, the automation is attached, which means that it is not editable. To edit the automation, you need to either make a copy or detach it.


You can make some changes to the automation settings, such as Propagation Labels, Run as, Roles, etc., without having to detach, or duplicate the automation.

While the automation is detached, it is not updated by the content pack. This may be useful when you want to update the automation without breaking customization. If you want to update the automation through content pack updates, you need to reattach it, but any changes are overwritten by the content pack on upgrade. If you want to keep the changes, make a copy before reattaching.

Automation Settings

In Script settings you can define the following information:





An identifying name for the automation.

Language type

Select the automation language.


The automation language version.


A meaningful description for the automation.


Predefined script identifiers that determine where the automation is available. For example, to use this automation as a pre-processing automation, it must be tagged with the pre-processing tag.

Run on

The Cortex XSOAR server, single engine, or Load-balancing group.


You need to create an engine for this field to appear.

Propagation Labels

(Multi-tenant) Select the propagation labels for the automation.


Whether the automation is available for playbook tasks, indicator types, incident types and fields.

Content Pack

The Content Pack for which the automation belongs.


Click Add Argument to add new arguments as required.




An identifying name.


Makes the argument mandatory.


Makes this argument the default argument for the automation.


Hides the argument from being displayed in the UI and in logs.


A meaningful description for the argument.

Initial value

The initial default value for the argument.

Is array

Specifies that the argument is an array.


The value type of the argument. Options are Unknown, Key-Value, Text Area.

List options

CSV list of argument values.


Click Add output to add the outputs according to string, number, date, boolean, etc. For more information, see Context and Outputs.



Context Path

A dot notation representation of the path to access the Context. For example, ThreatStream.Analysis.ReportID.


A short description of what the context path represents. For example, The ID of the report submitted to the sandbox.


The value type of the context path, such as string, number, etc. Enables Cortex XSOAR to format the data correctly.


Click Add important to add information for backwards compatibility only for previous Cortex XSOAR versions.




Password Protect

Enables you to add a password for the automation, which will be required when running the script from the CLI

Run as

Determines which role the automation runs as. Select the role from the dropdown list.


The assigned roles determine who can run the automation.




Timeout (seconds)

Time (in seconds) before the automation times out. Default is 180.

Run on

The Cortex XSOAR server.

Docker image name

For Python automation, the name of the Docker image to use to run this automation.

Run on a separate container

Runs the automation on a separate container.

Depends on Commands

You can set the commands that the automation depends on directly from these settings. You still have the option to set the dependencies in the automation YAML file.