Using Built-in Actions

Moveworks provides built-in actions for common platform operations — user lookups, notifications, approvals, and LLM generation. These are ready to use without creating connectors or HTTP actions.

When you want to use a built-in action in your Compound Action, use mw.{{action_name}} as your action_name. For example:

- action:
    action_name: mw.get_user_by_email
    output_key: user_result
    input_args:
      user_email: meta_info.user.email_addr

Action Reference

get_user_by_email

Retrieve a single user record by email address. Returns all user attributes including custom attributes.

Action Name: mw.get_user_by_email

Input Parameters:

Example:

- action:
    output_key: user_result
    action_name: mw.get_user_by_email
    input_args:
      user_email: meta_info.user.email_addr

Response Schema:

{
  "user": {
    "full_name": "John Doe",
    "email_addr": "john.doe@moveworks.ai",
    "record_id": "10769610738892443",
    "department": "Engineering",
    ...
  }
}

Reference: data.user_result.user.email_addr, data.user_result.user.full_name, etc.

You can find the full list of user fields in our User Attribute Reference.

batch_get_users_by_email

Retrieve multiple users by email in a single call. Use this when you need to resolve several users (e.g., an approver and a requester).

Action Name: mw.batch_get_users_by_email

Input Parameters:

Example:

- action:
    output_key: target_users
    action_name: mw.batch_get_users_by_email
    input_args:
      user_emails:
        - data.manager_email
        - data.employee_email

Response Schema:

{
  "user_records": [
    {
      "lookup_id": "12345",
      "user": {
        "full_name": "John Doe",
        "email_addr": "john.doe@moveworks.ai",
        "record_id": "10769610738892443",
        ...
      }
    },
    {
      "lookup_id": "67890",
      "user": {
        "full_name": "Jane Smith",
        "email_addr": "jane.smith@moveworks.ai",
        "record_id": "10769610738892444",
        ...
      }
    }
  ]
}

Reference: data.target_users.user_records[0].user for the first user, data.target_users.user_records[-1].user for the last.

You can find the full list of user fields in our User Attribute Reference.

create_generic_approval_request

Create a Moveworks in-bot approval request. The compound action pauses execution until the approval is answered (up to 30 days).

Action Name: mw.create_generic_approval_request

Input Parameters:

Example:

- action:
    output_key: approval_result
    action_name: mw.create_generic_approval_request
    input_args:
      title: '''Access to Coupa Request'''
      approval_details: '''Need access to Coupa for procurement workflow'''
      approvers:
        - data.manager_user.user
      users_requested_for:
        - data.requesting_user.user
    progress_updates:
      on_pending: Sending approval request...
      on_complete: Approval received.

Response Schema (Approved):

{
  "status": "APPROVED",
  "approved_by": [
    {
      "full_name": "John Doe",
      "record_id": "10769610738892443",
      "email_addr": "john@moveworks.ai",
      "comment": "Looks good, approved."
    }
  ],
  "state": {
    "initial_status": "PROCESSING",
    "status": {
      "john@moveworks.ai": "APPROVED"
    }
  }
}

Response Schema (Denied):

{
  "status": "DENIED",
  "denied_by": [
    {
      "full_name": "John Doe",
      "record_id": "10769610738892443",
      "email_addr": "john@moveworks.ai",
      "comment": "Budget not available for Q2."
    }
  ],
  "state": {
    "initial_status": "PROCESSING",
    "status": {
      "john@moveworks.ai": "DENIED"
    }
  }
}

Reference: data.approval_result.status returns "APPROVED" or "DENIED" (uppercase).

To access the approver/denier’s comment (e.g., a denial justification):

# Access the denial comment
message: data.approval_result.denied_by[0].comment
# Access an approval comment
message: data.approval_result.approved_by[0].comment

batch_send_plaintext_chat_notification

Send a chat notification to one or more users. Use with batch_get_users_by_email to resolve user record IDs first.

Action Name: mw.batch_send_plaintext_chat_notification

Input Parameters:

Example:

- action:
    output_key: target_users
    action_name: mw.batch_get_users_by_email
    input_args:
      user_emails:
        - '''alice@example.com'''
        - '''bob@example.com'''
- action:
    action_name: mw.batch_send_plaintext_chat_notification
    output_key: notification_result
    input_args:
      notifications:
        MAP():
          items: data.target_users.user_records
          converter:
            user_record_id: item.user.record_id
            message: '''Please complete your quarterly review.'''

Response Schema:

{
  "user_notification_tracker_details": [
    {
      "notification_tracker_uuid": "abc-123-def"
    }
  ]
}

generate_text_action

Generate free-form text output using an LLM. Use for summarization, classification, formatting, and other text generation tasks.

Action Name: mw.generate_text_action

Input Parameters:

Example:

- action:
    action_name: mw.generate_text_action
    output_key: summary_result
    input_args:
      system_prompt: '''Summarize the following text in 2-3 sentences.'''
      user_input: data.article_content
      model: '''gpt-5.4-mini'''

Response: data.summary_result.generated_output (string)

For full reference including supported models, see LLM Actions.

generate_structured_value_action

Generate structured JSON output matching a defined schema. Use for classification, entity extraction, and any task where you need typed output.

Action Name: mw.generate_structured_value_action

Input Parameters:

Example:

- action:
    action_name: mw.generate_structured_value_action
    output_key: classification_result
    input_args:
      payload: data.customer_review
      system_prompt: '''Analyze the sentiment of this review.'''
      output_schema: >-
        {
          "type": "object",
          "properties": {
            "sentiment": {"type": "string", "enum": ["positive", "negative", "neutral"]},
            "confidence": {"type": "number"}
          },
          "required": ["sentiment", "confidence"],
          "additionalProperties": false
        }

Response: data.classification_result.generated_output (object matching your schema)

For full reference including supported models, see LLM Actions.