Actions

App actions let you expose specific app functionality contextually within the Zoom app. They can appear in any Zoom product and help users execute actions, workflows, and automation within a product. App actions can also be triggered in Zoom AI Companion, which offers a more natural way to execute them.

Availability

  • Currently, app actions are only executable using API calls through connect endpoints. Support for webview and webhook-triggered actions is not yet available, but will be available in the future.
  • For now, app actions are available only for Zoom Virtual Agent (ZVA).

Prerequisites

You have created a Zoom general app configured with at least one connect endpoint. This is required to integrate third-party data into the Zoom platform.

Add action

To add an action, in your app's build page, select Actions and Triggers from the left navigation, and then click New action.

The wizard contains five overall steps to creating an action:

  1. Basic configuration to define the action.
  2. Where can users access it? to select the location in the Zoom client where users can use the action.
  3. Form Builder to specify the input fields defined in connect parameters that are available to this action.
  4. Transformation to map the incoming data and convert them from the API response values to the expected schema the Zoom application needs.
  5. Testing to verify your configuration is successful.

We explain each step further in the sections below.

Basic configuration

In the Basic configuration step, provide information to define the action.

  • What should happen when the user initiates this action - Currently only Custom actions is available. You can use custom actions to map developer-defined form fields, input/output mappings and definitions.
  • Action name - The display name of the action. This is the name that users see in the Zoom app for the action. Enter a descriptive, concise, action-oriented name that clearly describes what the action does.
  • Action ID - The unique identifier for the action. This is used to refer to the action within the app build flow. Enter a unique identifier for this action using snake_case format.
  • Description - An explanation of the action. Provide a clear, brief explanation of the action being configured. This info might be visible by users and used to trigger the right action in AI scenarios.
  • Action icon (Optional) - the icon to display with the action. Upload a square image in a file size less than 1MB. If no icon is uploaded, Zoom displays a default icon.
  • Action type - The main selection that defines how the action will be executed. Currently, the only supported action type is Trigger connect endpoint. In the Endpoint field, select the connect endpoint to use to execute the action.

Where you can users access it

In the Where you can users access it step, select the locations in the Zoom client you want the action to be available to users.

The available locations depend on the product you are using. For example, available locations for Chat would differ from locations for Meeting or Revenue Accelerator.

Form Builder

In the Form Builder step, you can add and configure input fields. Click Add field and select or enter values for:

  • Field ID - Enter an ID to use values of this field as a dynamic variable.
  • Field title - Enter a name for the field.
  • Description - A clear explanation of the field.
  • Data Type - Select whether the type is a string, number, integer, password.
  • Required - By default, all input parameters are marked as required. You can change this by toggling off the Required field.

You can use the Preview Form feature to preview the input form that users will see when they trigger this action.

Transformations

In the Transformations step, you map the incoming data and convert them from the API response values to the expected schema the Zoom application needs.

Select the endpoint you configured in Connect, then provide the following information:

  • Input mapping - Defines how to transform custom form inputs collected from end users into API inputs.
  • Output mapping - Defines how to transform API outputs (optional; if not provided, the raw API output becomes the final output).
  • Output definition - Defines the finalized output structure, including display names shown to end users and field specifications that conform to the expected object format.

To reference a custom form input value, use {{ctx.inputData.<fieldId>}}.

In addition to action-specific custom form inputs, you can access app-level or connection-level custom form values captured during app installation, using {{ctx.customField.<fieldId>}}.

Script mode

You can choose to use a user-friendly UI mode for simpler actions that only need to call a single API without complex data transformation logic, the UI mapping tool provides a faster setup process with its intuitive interface.

We recommend using Script mode to a fully customize your data to configure the action. It allows you to define an action that requires multiple API calls to complete a task. And gives you the freedom to write custom code and functions that can chain API calls together or manipulate data between steps. In script mode, use zm.fetch to invoke third-party endpoints.

For more information about zm.fetch, see Actions/zm.fetch on GitHub. It includes information on syntax, how to populate headers, data, and parameters using different structures, and complete examples with sample endpoints.

Test the action

In the Test step, you can verify your setup.

Select a connection and provide sample values for the custom form inputs. If the results meet expectations, you can save and finalize the action setup.

Both live and test connections are supported. To use a live connection, you must first install the app. To use a test connection, you must first configure it in the Connect section.