Quickstart #7: Scheduled Trigger Ambient Agent

Learning Objectives

In this Quickstart, you’ll build an Ambient Agent that is triggered on a schedule and notifies you with a weekly risk analysis of purchase requisitions. Along the way you’ll learn how to:

look up a target user(s) with built in mw.get_user_by_email
Use built in mw.generate_text_action and the Notify expression.
Configure a Scheduled Trigger to run a Compound Action on a defined cadence

Prerequisites

Access to the “Agent Studio” app.

What You’ll Build

The plugin you’ll build solves a common, practical problem: proactively surfacing valuable insights to the right person on a reliable cadence.

Before starting to build, we always recommend that you first map out the desired end user experience and the overall building blocks of an Ambient Agent.

Here is an example of the end-user experience we hope to enable:

To get to this experience, the Ambient Agent will consist of one Compound Action

You’ll build an end-to-end working Ambient Agent plugin via these phases:

Set up an action.
Build the Compound Action, containing three actions and a notify expression.
Add the Compound Action to your AI Agent by launching a Plugin.

Let’s get started!

Phase 0: Set Up Your Connector

You may be able to skip this phase if you’ve completed another Agent Studio Quickstart guide. You can skip this phase if:

You have already created a firstname_lastname_moveworks_purple connector. This section will walk you through how to set up your own connector to the Moveworks Purple API (which powers the Actions used in the Quickstart guides), so no need to repeat this step if you’ve done it before.

In this guide, the purchase requisition data that your Action will access lives behind the Moveworks Purple API Connector. To access this data via your Action, you need to set up a (reusable) connector:

Open HTTP Connectors

Navigate to the App Picker in the top right corner of Agent Studio. Click on “HTTP Connectors”, which will take you to another browser tab.

Create a new connector

Click “Create” in the top right corner. Fill out the following information for your Connector (be sure to replace “firstname” and “lastname” with your corresponding information).

Field	Value to enter (replace “firstname” & “lastname” with your info)
Connection Name	`firstname_lastname_moveworks_purple`
Description	`Moveworks Purple APIs Connector`
Base Url	`https://marketplace.moveworks.com`
Auth Config	`No Auth`

Save your connector

Click “Save” on the bottom right corner.

You’re now fully ready to start building. On to Phase 1!

Phase 1: Create the HTTP Action to fetch purchase requisitions

Navigate to Agent Studio

Navigate to the App Picker in the top right corner. Click on “Agent Studio”.

Create a new HTTP Action

Navigate to a new HTTP Action. Click “Create”.

Set the title and description

Set the following Title and Description for your Action (be sure to replace “firstname” and “lastname” with your corresponding information).

Field	Value to enter (replace “first” & “last” with your info)
Title	`firstname_lastname_get_purchase_requisitions_action`
Description	`Retrieves all purchase requistions`

Import the API details

Enter the details of your API:

Click on the “Import” icon to the right of the “TEST” button.
Paste the following cURL command (this represents an endpoint which has one dynamic user email query parameter):

1 curl -X GET "https://marketplace.moveworks.com/api/purple-suite/erp/purchase-requests/history"

Click “Import” (your Action should now be auto-populated with details).

Add your connector

Add your Connector:

Navigate to the Connector tab.
Select the “Inherit from existing connector” radio button.
Select the Connector you created in Phase 0 from the dropdown (e.g. firstname_lastname_moveworks_purple).

Test your action

Test your Action:

Click on the “TEST” icon (to the right of the endpoint URL definition).
Inspect the Console to verify that you receive a successful response:

Publish

Click “Publish”.

Phase 2: Create Your Compound Action

You will now create a compound action to handle all the logic of the plugin when triggered by its schedule.

In this phase you will build a compound action with the action you configured to retrieve all purchase requistions, a built in action to lookup the target user’s Moveworks profile, and a Notify key to message provide the user the option to retrieve the insights if they select yes it will call a built in action to generate the risk analysis against the purchase requisition data

Create a new Compound Action

Navigate to a new “Compound Action” in the left nav and then click “Create”.

Set the title and description

Set the following title and description for your Compound Action (be sure to replace “first” and “last” with your corresponding information).

Field	Value
Title	`quickstart_first_last_purchase_requisition_insights_ca`
Description	`Sends a message to a user to get insights on purchase requisitions`

Configure the Compound Action expression

Copy the following expression into the Compound Action Editor and:

Replace <your email here> with your actual email on line 9
Replace <your pr action name here> on line 3 with whatever you named the action firstname_lastname_get_purchase_requisitions_action from step 1.

1 steps:
2   - action:
3       action_name: <your pr action name here>
4       output_key: purchase_requistions
5   - action:
6       action_name: mw.get_user_by_email
7       output_key: user_object
8       input_args:
9         user_email: '"<your email here>"'
10   - notify:
11       output_key: notify
12       recipient_id: data.user_object.user.id
13       message: '"Select yes if you would like to receive an analysis on purchase requisitions for the day"'
14       actions:
15         - key: yes_option
16           label: '"Yes"'
17           system_action:
18             action_name: mw.generate_text_action
19             input_args:
20               model: '"gpt-4.1"'
21               system_prompt: '"Analyze the purchase requistions for the day and risk based on item, dollar amount, and quantity. Callout any specific items that seem over priced or innacurate"'
22               user_input: data.purchase_requistions.$STRINGIFY_JSON()

Understanding the Compound Action

This Compound Action executes four actions in sequence, with the output of one step feeding into the next:

firstname_lastname_get_purchase_requisitions_action
1. Purpose: To identify the notification recipient.
2. Process: Custom made action you configured in step 1 to call the purple api and retrieve all purchase requisitions.
mw.get_user_by_email (Built-in Action)
1. Purpose: To identify the notification recipient.
2. Process: It takes a user’s email address (you defined statically) as its input and retrieves that user’s unique Moveworks profile ID.
mw.generate_text_action (Built-in Action)
1. Purpose: To create a polished message for the user and analyze for insights and callout risk.
2. Process: This action uses an LLM to interpret the raw data in the webhook payload and generate a structured, readable message.
notify (Compound Action Expression)
1. Purpose: To send the final notification.
2. Process: This action takes two inputs from the previous steps:
  1. Recipient ID: It uses the user ID that was retrieved by the mw.get_user_by_email action.
  2. Message: It uses the formatted text that was generated by the mw.generate_text_action action.

Publish

Click “Publish”.

All done with Actions!

You’re ready to move on — time to create your Plugin and put everything you’ve built together.

Phase 3: Configure and Launch Your Plugin

This section guides you through adding your Scheduled Trigger and Compound Action to a Plugin. A key step will be configuring the data flow, where you’ll specify exactly which fields from the webhook you want to forward to your action.

Create a new Plugin

Navigate to a new Plugin (navigate to the library and click “Create”).

Set the title and description

Set the following title and description for your Plugin (be sure to replace “first” and “last” with your corresponding information).

Field	Value to enter
Title	`quickstart_first_last_purchase_requisition_insights_notification`
Description	`Sends a message to a user to opt in to recieving purchase requisition insights.`

Configure the trigger and body

In the “Workflow Overview” section click “System” in the Trigger box.

Configure your system trigger in the slide-out pop up with:

Field Label	Value to enter/select
Select Trigger Type	`Schedule`
Schedule Type	`Interval`
Interval Schedule	`5 Minutes` (or whatever you want for testing) Minimum 1 minute*
Schedule Settings	Start: `Start Immediately after published`	Stop: After 1 occurence(s) (recommended to stop so it doesn’t keep notifying you)

Click on the “Body” in the “Workflow Overview
Configure your system body to connect your compound action to the plugin

Field label Value to enter/select
Select Type System body
Select an Action quickstart_first_last_purchase_requisition_insights_ca (Select yours)

Field label	Value to enter/select
Select Type	`System body`
Select an Action	`quickstart_first_last_purchase_requisition_insights_ca` (Select yours)

Publish your Plugin

Click “Publish” to launch your Plugin.

And just like that — you've finished building your Plugin, and now it's ready to use!

Congrats! Your plugin is complete. You should get notified from your AI assistant with a message including a button to receive insights.

Reflecting on this Plugin

The plugin you just created uses several key Agent Studio concepts, from Schedule Triggers to LLM actions to Compound Actions. You’ve successfully built a plugin that:

Runs on a cadence on it’s own
Grabs all the purchase requisitions.
Processes that data to find insights onto purchase requisitions happening daily.

Awesome work sticking with it!

Next step: Test your plugin. See our Testing & Error Handling guide for how to test, debug, and handle errors in production.

1	steps:
2	- action:
3	action_name: <your pr action name here>
4	output_key: purchase_requistions
5	- action:
6	action_name: mw.get_user_by_email
7	output_key: user_object
8	input_args:
9	user_email: '"<your email here>"'
10	- notify:
11	output_key: notify
12	recipient_id: data.user_object.user.id
13	message: '"Select yes if you would like to receive an analysis on purchase requisitions for the day"'
14	actions:
15	- key: yes_option
16	label: '"Yes"'
17	system_action:
18	action_name: mw.generate_text_action
19	input_args:
20	model: '"gpt-4.1"'
21	system_prompt: '"Analyze the purchase requistions for the day and risk based on item, dollar amount, and quantity. Callout any specific items that seem over priced or innacurate"'
22	user_input: data.purchase_requistions.$STRINGIFY_JSON()