In the Lab/What's New

What's New — September 11, 2026

The LitmusLab Team

The LitmusLab Team

September 11, 2026·4 min read

Week of September 5–11, 2026


Multiple UI Monitors per project 🔥

A project can now run more than one UI Monitor, each watching its own slice of your site. Point one monitor at your marketing pages and another at your authenticated app, each with its own test suites to match against, its own sampling rate, and its own on/off switch. All of it runs from a single embed snippet installed once, site-wide. No more marketing traffic showing up as noise against app-only Flows, and no more choosing between monitoring everything or nothing.

Privacy Controls for session capture

Every project now protects passwords, and anything that looks like a card number, CVV, or Social Security number, automatically. Nothing to configure. On top of that floor, add your own Block, Ignore, and Mask rules from the Capture tab to protect anything specific to your app. Even screenshots generated later, for UI Monitor checks or the Graph tab, carry the same redaction forward, so a protected field never becomes visible some other way.

More control over what gets captured

The hosted capture script now exposes a small client API: pause recording through a sensitive flow, attach your own session properties, identify a visitor, or opt someone out, all from your own code. And your own scripted test runs are excluded from your live-traffic data by default: LitmusLab's test runner tags its own requests so they never show up looking like a real session. Add your own pattern list too, if you want to exclude other automated traffic.

Sharper, more trustworthy results

Check results now show you the actual value being compared, not the raw {{VARIABLE}} placeholder, so you can see exactly what was checked at a glance. That includes checks that reference environment-specific values like your app's URL, which now resolve correctly for live sessions the same way they already did for test runs. Reference screenshots on the Graph tab stay current as your test suites evolve instead of getting stuck on a stale example. And matching now tells apart different outcomes of the same page (a failed login and a successful one no longer get treated as the same state), so a check failure means what it says.