Settings

Crawl settings

Max pages limits how many pages a discovery crawl can find. Keep this sensible. A smaller set is easier to review and cheaper to run.

Max depth controls how far link crawling can travel from the base URL.

Recrawl interval controls how often viz looks for added or removed pages. Turn it off if you only want manual discovery.

Respect robots decides whether the crawler follows robots.txt rules. Keep this enabled unless you own the site and have a clear reason to turn it off.

Screenshot settings

Viewports define the screen sizes that are captured. Start with the smallest useful set, for example mobile and desktop.

Wait strategy decides when viz takes the screenshot:

• networkidle waits for network activity to settle.
• domcontentloaded waits for the HTML document to be ready.
• load waits for the page load event.

Wait time adds a fixed delay before capture. Use it for animations or delayed content, but avoid large values unless needed.

Full page captures the whole scrollable page. Turn it off if you only care about the first viewport.

Custom CSS

Custom CSS is injected just before the screenshot. Use it to hide noisy elements:

.chat-widget,
.live-price,
[data-testid="timestamp"] {
  visibility: hidden !important;
}

Prefer hiding dynamic elements over approving the same false-positive diff again and again.

Comparison settings

Threshold controls how sensitive pixel comparison is.

Failure threshold controls how much of the page can differ before viz reports a visual change.

Keep these settings strict enough to catch real regressions, but relaxed enough to avoid tiny rendering noise.