|
Trigger For
|
- **Form** - Lens action is used when ServiceNow AI Lens is triggered to auto-fill a form. Apply a transformation logic to the Lens response before the form is auto-filled by using the [https://raw.githubusercontent.com/ServiceNow/ServiceNowDocs/australia/markdown/australia/intelligent-experiences/servicenow-lens/servicenow-lens-actions.md](https://raw.githubusercontent.com/ServiceNow/ServiceNowDocs/australia/markdown/australia/intelligent-experiences/servicenow-lens/servicenow-lens-actions.md) option.
This option is available when you select **Desktop** or **Instance** from the Trigger From list.
- **Service** - Lens action is used when ServiceNow AI Lens is used as a service. Select this option to trigger Lens from an instance to auto-populate a **Workspace form**, or to trigger it using Virtual Agent or Script Include.
This option is available when you select **Instance** from the Trigger From list.
- **Others** - Lens action is used when ServiceNow AI Lens is triggered to display a preview of extracted data and perform post processing steps, if applicable.
This option is available when you select **Desktop** from the Trigger From list.
- **Excel Mapping** -
Option to specify that you can use this Lens action to attach an Excel sheet and let ServiceNow AI Lens auto-map its column headers with the columns of a table. After the auto-mapping is done, you can view and change the mapping, and insert the Excel sheet data to the table.
This option is available when you select **Desktop** from the Trigger From list.
|
|
Order
|
Execution priority of the Lens action for a given table:- **1**: Highest priority
- **999999**: Least priority
The default value is set to 100.
For example, if any user is assigned to two Lens action records for the same table, the Lens action with the least number is used during Lens execution.
|
|
Table
|
Table to which this Lens action applies. Selected table provides context during Lens execution when specific fields are selected or script is provided.For example, Incident table.
|
|
Request Context
|
Option to provide enhanced context during Lens execution by selecting specific fields or providing custom schema via script.- **Supported Fields** - Selected fields for providing enhanced context for Lens execution.
- **Advanced** - Script to generate a custom schema object for providing enhanced context for Lens execution.
|
|
Fields
|
Table fields that you can select to provide enhanced context for Lens execution. Select fields to help Lens determine what information to extract from images. Only fields supported by Lens are available for selection. For more information, see [Field types supported](https://raw.githubusercontent.com/ServiceNow/ServiceNowDocs/australia/markdown/australia/intelligent-experiences/servicenow-lens/field-types-supported.md).This field appears only when **Supported Fields** is selected from the **Request Context** choice field.
|
|
Pre Processing Script
|
Script to generate a custom schema object to provide enhanced context for Lens execution. Custom schema help Lens determine what information to extract from images.This field appears only when **Advanced** is selected from the **Request Context** choice field.
Example sample script:
```
function preProcessRequest(lensActionRecord, inputJSON) {
// Initialize the schema object
var schemaObj = {};
/*
This is a utility method to generate schema object,
users can pass tableName and formFields as an array of fields for the given table to generate custom schema that is passed to the LLM.
//Return schemaObj adhering to the structure defined in example
// tableName = 'incident'
tableName = inputJSON.tableName;
//formFields - array of fields for the given table.
//formFields = ['short_description', 'comments', 'impact'];
fieldsToProcess = inputJSON.formFields;
schemaObj = new global.AILensSchemaUtil().generateSchema(tableName, fieldsToProcess);
*/
return schemaObj;
}
var schemaObj = preProcessRequest(lensActionRecord, inputJSON);
```
Parameters accepted by the `preProcessRequest` method:
- **lensActionRecord** \(Input\) - GlideRecord of the Lens action that is executed.
- **inputJSON** \(Input\) - Optional JSON object that users can pass through service APIs.
- **schemaObj** \(Output\) - Returned JSON object of custom schema.
Example of the returned schemaObj object:
```
schemaObj = {
"schemaPayload": {
"type": "json_schema",
"json_schema": {
"name": "schema_extraction",
"schema": {
"type": "object",
"properties": {
"short_description": {
"type": "string",
"label": "Short description"
},
"comments": {
"type": "string",
"label": "Additional comments"
},
"impact": {
"type": "choice",
"label": "Impact",
"choiceValues": ["1", "2", "3"],
"choiceLabels": {
"1 - High": "1",
"2 - Medium": "2",
"3 - Low": "3"
}
}
},
"required": [
"short_description",
"comments",
"impact"
],
"additionalProperties": false
}
}
},
"fieldLabels": ["short_description", "comments", "impact"] //Optional key
}
```
|
|
Default Additional Directions
|
Default instructions that are supplied to AI in addition to the user-provided instructions.For example, `Summarize this paper by extracting the title, author name, publication date in the dd-mm-yyyy format, abstract, and keywords`
|
|
Transform Response
|
Option to specify the transformation script for transforming the Lens response before auto-filling the form.This field appears only when **Form** is selected from the **Trigger For** choice field.
|
|
Transformation Script
|
Script that transforms the Lens response before populating the form.This field appears only when the **Transform Response** option is selected.
Sample script:
```
function transformResponseScript(lensActionRecord, lensResponse) {
// Initialize the schema object
var transformResponse = {};
//Add your code here
//Return schemaObj adhering to the structure defined in example
return transformResponse;
}
var transformResponse = transformResponseScript(lensActionRecord, lensResponse);
```
Parameters accepted by the `transformResponseScript` method:
- **lensActionRecord** \(Input\) - GlideRecord of the Lens action that is executed.
- **lensResponse** \(Input\) - JSON object of the Lens response provided after Lens execution.
- **transformResponse** \(Output\) - Returned JSON object of transformed schema.
Example of the returned transformResponse object:
```
transformResponse = {
"short_description": "Service Degradation Error in Order Processing System",
"description": "The Order Processing API v2.1 encountered a service degradation issue in the Production environment",
"urgency": "1"
}
```
|
|
Enable Post Processing
|
Option to specify advanced processing instructions after Lens execution is complete.This field appears only when **Others** or **Service** is selected from the **Trigger For** field.
|
|
Post Processing Script
|
Script that is run asynchronously after Lens execution to further process the Lens response, such as triggering a flow, subflow, actions, or AI Agent.This field appears only when the **Enable Post Processing** option is selected.
```
function postProcessResponse(lensActionRecord, lensResponse, lensExecutionRecord) {
//Add your code here
}
postProcessResponse(lensActionRecord, lensResponse, lensExecutionRecord);
```
Parameters accepted by the `postProcessResponse` method:
- **lensActionRecord** \(Input\) - GlideRecord of the Lens action that is executed.
- **lensResponse** \(Input\) - JSON object of the Lens response provided after Lens execution.
- **lensExecutionRecord** \(Input\) - GlideRecord of the Lens execution record.
|
|
Wait for processed response
|
Option to wait for ServiceNow AI Lens to display the output of post-processing on the Preview window. If enabled, it waits for the post processing output. Else, it doesn't wait for post processing output but post processing continues in the background.
|
|
Max wait time \(seconds\)
|
The maximum time ServiceNow AI Lens waits to display the output of post-processing. If the output doesn't appear within this duration, the session times out while the post-processing continues in the background until completion.
|
|
Attach Captured Images To Record
|
Option to attach the captured images to the record that is auto-filled using ServiceNow AI Lens.This option is enabled by default.
|