Build and edit a monitor

A monitor’s Steps define what to capture and read. It starts with one Open page step; add Read steps to collect values from different areas of that page. All Read steps use the same capture.

Select a monitor and open Steps, or follow Your first monitor to create one.

Configure the page capture

Enter the full https:// or http:// URL in Open page, then choose Test step to capture it using your selected Chrome profile. Check that the screenshot shows the content you need.

Open the card’s gear button to adjust:

Screenshot delay: how long to wait after the page opens before capturing it. Increase this if content is still loading. It accepts 0–60 seconds.
Device viewport: the width and height used to lay out the page, its pixel density (Device scale factor), and Mobile layout behavior. Resizing your Chrome window does not change these settings.

After changing capture settings, test Open page again and check your selected regions. Their positions do not automatically adjust when the page layout changes.

Add readings

Drag a rectangle around the information you need in the screenshot. Include nearby labels that help identify the values.
Choose Add step → Read below the screenshot.
Give each field a descriptive name and choose its type. Use Add field to collect several values from the same region.

Type	Use for	Example
Number	Prices, counts, or other numeric values	Price: `18.50`
Text	Labels, statuses, or other text	Availability: `In stock`

Use Reading instructions (optional) to resolve ambiguity. For a Price field, try “Read the current purchase price, excluding shipping.” For Availability, try “Read the stock status for the selected size.” Leave instructions empty when the name and region are sufficient.

Add another Read step when the information is in a different area. Each Read step sends its selected screenshot region, relevant page content, and field instructions to its model. A step reports an error if any requested field is missing, uncertain, or invalid.

Open the Read card’s gear button to choose its AI model and Timeout. New steps use the app’s default model and a 60-second timeout. Changing the default model later does not change existing steps. See Set up AI models for setup details.

Adjust regions and steps

Click a Read card to select its region, then move or resize the rectangle on the source screenshot. Use the screenshot’s zoom controls for precise adjustments.

Drag Read cards or use their up/down controls to change their order. Use a card’s delete button to remove it. Open page always stays first and cannot be deleted.

Editing a Read step or its region clears its test readings and those of later steps. Test the affected steps again. Previously saved checks remain in History.

Test, save, and run

Action	What it does
Test step	Tests your current edits. Runs unfinished earlier steps first, then the selected step. Results are temporary and do not enter History.
Save (⌘S)	Saves changes to Steps, Schedule, and the monitor name together. It does not save screenshots or test results.
Discard changes	Restores the saved Steps, Schedule, and monitor name after confirmation.
Check now	Captures the page again, runs the saved workflow, and stores the outcome in History. Save or discard all pending edits first.

A Read test can reuse the existing screenshot. To test against the current webpage, test Open page first.

Before saving, compare test readings with the screenshot. After saving, run Check now and verify the result in History. Hover over a disabled control for its explanation.

Scheduled checks continue to use saved settings while you edit. Unsaved changes take effect only after Save; a check already running keeps the settings it started with. See Schedule checks to set up automatic collection.

Rename or delete a monitor

Open Details to edit Monitor name, then save. To remove the monitor, choose Delete monitor… and confirm. This permanently removes its steps, schedule, check history, and readings.