Screen, anchor, and step properties in AI Desktop Actions

  • Release version: Australia
  • Updated May 25, 2026
  • 3 minutes to read
    • Summarize
    Summarized using AI
    This content was generated using new OpenAI-powered functionality. Results are provided on an as is basis and are not guaranteed to be accurate or complete.

    Summary of Screen, anchor, and step properties in AI Desktop Actions

    This document explains the properties of screens, anchors, and steps used in AI Desktop Actions within ServiceNow’s AI Agent Studio. These properties allow you to customize how automation interacts with captured screens and UI elements, controlling execution order, interaction details, and recognition accuracy to optimize automation flows.

    Show full answer Show less

    Screen Properties

    • Name: Unique identifier for each captured screen, visible in the Screens and steps panel.
    • Execution order: Defines the sequence in which multiple screens execute; adjustable by dragging screens in the panel.

    Anchor Properties

    Anchors represent specific regions on a screen and have properties that control their location, size, and behavior during automation.

    • Name: Unique identifier for the anchor.
    • Execution order: Sequence of anchor execution when multiple anchors exist.
    • X, Y, Width, Height: Define the position and size of the anchor area on the screen.
    • Convert to grayscale: Option to convert the captured screen area to grayscale for image processing (default is False).
    • Search in Scale: Allows scaling the target image incrementally between 100% and 150% to improve matching.
    • Confidence threshold: Sets the required accuracy for image matching before executing steps (default 95%).
    • Retry with scroll: Option to scroll the target image to find a match if initial attempt fails.
    • Wait for image: Determines if automation waits for the image to appear before proceeding (default True).
    • Max wait time: Maximum seconds to wait for the target image or application to appear.
    • Delay before/after: Specifies wait times before and after executing the anchor, controlling pacing of automation.

    Step Properties

    Steps define actions performed on anchors and have common properties across all step types, plus specific properties depending on the action.

    Common Step Properties

    • Name: Unique name for the step.
    • Execution order: Order of step execution within an anchor, adjustable by dragging.
    • Type: Specifies the type of action (e.g., Set Text, Click, Get Text).
    • Description: Optional descriptive note for the step.
    • X, Y: Coordinates of the interaction point relative to the anchor.
    • Delay before/after: Seconds to wait before or after executing the step.

    Set Text Step

    • Use parameter: Option to input text dynamically from mapped Desktop action parameters.
    • Static value: Fixed text to enter when not using parameters.

    Mouse Click Step

    • Mouse click type: Specifies the mouse action such as left/right click, double click, middle click, drag, drop, scroll, or paste.

    Send Keys Step

    • Clear existing value: Option to clear the field before entering new text.
    • Use parameter: Enables sending text from a Desktop action parameter.
    • Static value: Text entered if parameter use is not selected.
    • Focus with mouse click: Option to set focus on the target area before sending keys.

    OCR Read Text Step

    • Width and Height: Define the area on the image to extract text from using OCR.

    Practical Use for ServiceNow Customers

    Understanding and configuring these properties enables precise control over UI automation steps in AI Desktop Actions. By setting execution order, timing delays, and interaction types, you can tailor automations to reliably interact with applications, handle dynamic content with image matching confidence and scaling, and extract or input data efficiently. This flexibility improves automation accuracy, reduces errors, and enhances the overall effectiveness of AI-driven desktop workflows integrated with ServiceNow.

    Learn about the properties of screens, anchors, and steps. There are multiple types of steps and each step type has distinct properties. You can update the properties to modify the behavior of the steps.

    Screen

    Table 1. Screen properties
    Property Description
    Name Unique name for the captured screen. The updated name reflects in the Screens and steps panel.
    Execution order Order in which the screen is executed when multiple screens are captured. You can drag the screen at the desired order in the Screens and steps panel.

    Anchor

    Table 2. Anchor properties
    Property Description
    Name Unique name of the added anchor. The updated name reflects in the Screens and steps panel.
    Execution order Order in which the anchor is executed when multiple anchors are added.
    X Location of the top-left corner of the anchor on the screen along the X-axis.
    Y Location of the top-left corner of the anchor on the screen along the Y-axis.
    Width Width of the area highlighted as anchor.
    Height Height of the area highlighted as anchor.
    Convert to grayscale Option to convert the captured screen as a grayscale image. The default value is False.
    Search in Scale Option to specify whether the component scales the target image before trying to match the source image. The scaling is done in an increment of 10% each between 100% and 150%. The default value is False.
    Confidence threshold Specifies the accuracy in matching the image captured before the component performs a step. The value of 1 defines a 100% match while 0.5 defines 50% match. The default value is 0.95 or 95% match.
    Retry with scroll Option to scroll the target image before trying to match the source image.
    Wait for image Option to specify if the automation must wait for the image to appear on the screen. The default value is True.
    Max wait time Number of seconds to wait for the image or application to appear on the screen.
    Delay before Delay in number of seconds before the anchor is executed. The automation waits until the specified time elapses before executing the anchor.
    Delay after Delay in number of seconds after the anchor is executed. The automation will wait until the specified time elapses before executing the next anchor.

    Step

    Common properties for steps

    The following properties apply to all steps - Set Text, Click, Mouse Click, Send Keys, Get Text, Get Table, and OCR Read Text.

    Table 3. Common properties
    Property Description
    Name Unique name of the step. The updated name reflects in the Screens and steps panel.
    Execution order Order in which this step is executed when multiple steps are set up. This order is assigned when you add a step from the anchor. You can drag the step under its anchor at the desired order in the Screens and Steps panel.
    Type Type of the step, such as Set Text, Get Text, Mouse Click.
    Description Helpful description of this step.
    X Location of the green plus icon along the X-axis relative to the anchor.
    Y Location of the green plus icon along the Y-axis relative to the anchor.
    Delay before Delay in seconds before the action is executed.
    Delay after Delay in seconds after the action is executed.
    Set Text
    Table 4. Set Text properties
    Property Description
    Use parameter Option to supply the value from a Desktop action parameter record mapped in AI Agent Studio. When selected, the Static value field is not used.
    Static value The text to enter in the field when Use parameter is not selected.
    Mouse Click
    Table 5. Mouse Click properties
    Property Description
    Mouse click type Type of mouse click action to set for this step:
    • Left click
    • Left double click
    • Middle click
    • Right click
    • Right double click
    • Move
    • Drag
    • Drop
    • Scroll down
    • Scroll up
    • Paste
    Send Keys
    Table 6. Send Keys properties
    Property Description
    Clear existing value Option to specify if the step clears the existing value on a field before setting the text.
    Use parameter Option to supply the value from a Desktop action parameter record mapped in AI Agent Studio. When selected, the Static value field is not used.
    Static value The text to enter in the field when Use parameter is not selected.
    Focus with mouse click Option to use the mouse click to focus on the area in the application where the mouse inputs are passed.
    OCR Read Text
    Table 7. OCR Read Text properties
    Property Description
    Width Width of the area in image to recognize and get the text.
    Height Height of the area in image to recognize and get the text.