March'26 Updates: Browser Cloud, KaneAI Conditional Logic, iOS Offline Mode & More

Browser Cloud launched on TestMu AI in March. KaneAI picked up four new capabilities, real device testing expanded for iOS and Espresso, and HyperExecute, SmartUI, Test Manager, and Test Insights each shipped improvements.

Here's what landed.

Introducing Browser Cloud on TestMu AI

Spinning up browser environments for testing, scraping, or AI agent workflows no longer requires infrastructure overhead. Browser Cloud on TestMu AI gives you instant access to 3,000+ browser and OS combinations with sub-second startup and sessions that run up to 24 hours.

• Framework-Agnostic SDK: Connect with Playwright, Puppeteer, Selenium, or your custom AI agent integrations. Spin up hundreds of headless browsers on demand via a simple SDK call.
• Session Persistence: Cookies, local storage, and login state carry across sessions. No re-authentication, no lost context.
• Geo-Targeted Proxies: Run your agents across 180+ geolocations. Every request looks local, so you can validate geo-specific content and pricing.
• Built-In Stealth: Advanced techniques to prevent bot detection, so your automation runs without interruptions.
• Full Observability: Live interactions, video replays, network and console logs, and step-by-step command replays to debug failed sessions fast.
• Local Testing Tunnel: Access apps behind firewalls or VPNs without exposing them to the public internet.

Get started with the Browser Cloud documentation.

New Features in KaneAI

Four updates shipped in KaneAI this month, covering element validation, global variables, conditional logic, and authoring flow improvements.

Element State and Attribute Assertions

No more writing custom code just to check if a button is enabled or a dropdown is selected. KaneAI now lets you assert element states and attributes using plain language.

Describe what you expect, like "verify the submit button is enabled" or "check the input field has value '[email protected]'", and KaneAI handles the validation for you.

Read the element state and attribute assertions guide.

TOTP Variables, Conditional Logic, and Slash Commands

• TOTP Variables Are Now Global: TOTP variables have moved from local to global scope. You can now access them across all your tests without redefining them each time. Before your next edit, check the guide on migrating local TOTP variables.
• If-Else-If and Conditional Modules: You can now use if-else-if blocks in the authoring flow and add modules within them to build complete conditional test logic, no workarounds needed. Learn more about conditional logic in KaneAI.
• Slash Commands in Add Step Flow: When inserting a step between existing steps, you can now use / slash commands to add Modules, API, JS, and DB steps alongside natural language input. Faster step insertion, same familiar flow.

Latest Enhancements in App Automation

Three updates for mobile and app testing this month, covering network simulation, mock backends, and OCR-based interactions.

iOS Offline Mode

Simulate a complete network disconnection on real iOS devices. Test cached data, offline fallbacks, and reconnection behavior. Works with KaneAI, manual testing, and Appium. Android offline support in KaneAI is coming soon.

See the docs for network throttling in KaneAI and offline mode for Appium.

Mock Server Support for Espresso

If your Espresso tests rely on MockWebServer or localhost mock servers, you can now run them on TestMu AI real devices. Control exactly what your backend returns and test error handling, edge cases, and timeout scenarios.

Here's how to run Espresso with MockWebServer on real devices.

Button Click By Text (OCR)

Not every UI element has a stable locator. Buttons with dynamic IDs, custom controls, or icon-only actions can stall your automation and add maintenance overhead.

Button Click By Text uses OCR to interact with elements based on the text you actually see on screen. If your users can read it and tap it, your tests can too. Available on both real and virtual devices.

Learn more about Button Click By Text.

Instrumentation Logs for XCUI and Espresso on HyperExecute

Debugging a failed XCUI or Espresso run on HyperExecute no longer means digging through raw device logs. Instrumentation logs are now surfaced directly, giving you clear visibility into setup, teardown, and device-level interactions.

When a test fails, you can pinpoint whether the issue was in test setup, a specific assertion, or device teardown, without switching between multiple log sources.

Refer to the HyperExecute YAML configuration guide.

Custom Viewport Support in SmartUI CLI

Default breakpoints don't always match what your users actually see. You can now define custom viewports directly in your SmartUI JSON config file.

Set the exact screen dimensions that matter to your product and capture visual snapshots at those resolutions. Your visual regression tests now reflect real usage, not just standard breakpoints.

Learn more about SmartUI SDK configuration options.

Folder-Based Test Organization in Test Manager

Scrolling through a flat list of test instances to find what failed is no longer necessary. Test instances in the Test Run Details page are now grouped by their original folder structure, so your runs mirror how you actually organize your tests.

You also get folder-level status rollups. One glance tells you which areas passed, which failed, and where to focus next.

Check out the test run management guide.

Insights API and Expanded RCA Log Coverage in Test Insights

Your test data should work harder for you. Here's what's new in TestMu AI Test Insights.

• Insights API: Pull test metrics into your own dashboards or plug test intelligence into your CI pipeline. The Insights API lets you access your data programmatically and integrate it wherever you need. Explore the Insights API documentation.
• RCA Log Coverage Expansion: If you run Cypress, Detox, or Maestro, your AI Root Cause Analysis just got sharper. RCA now ingests command logs, runner logs, flow execution logs, and device interaction logs for these frameworks, bringing them on par with Selenium, Appium, and Playwright. See the AI Root Cause Analysis guide.

Summing Up

March brought meaningful updates across every area of the platform. The Browser Cloud launch opens up new workflows for browser-based agents and scraping. KaneAI now supports conditional logic, global TOTP variables, and plain-language element assertions. Real device testing expanded with iOS offline mode and Espresso mock server support.

HyperExecute surfaces instrumentation logs for XCUI and Espresso directly. SmartUI supports custom viewports. Test Manager groups test instances by folder. And Test Insights opens up programmatic access through the Insights API with broader RCA log coverage.