As part of our ongoing efforts to improve agentic test authoring, we’ve made several updates that improve the precision and coverage of generated tests.
More precise assertions and element selection
The mabl agent now has visibility into the page’s DOM structure, which means it can make smarter decisions about how to identify elements and generate selectors. Combined with support for ARIA labels as a selector strategy, the agent produces more stable, accessibility-aware tests that better reflect how users interact with the page.
We’ve also improved assertion generation so the agent targets your application’s critical validations more precisely, helping you catch real issues faster while reducing test maintenance.
Support for multi-tab testing
The agent can now generate steps that switch between browser tabs. This expands agent coverage to multi-tab workflows such as OAuth popups, payment processing flows, and actions that open new windows. Previously, multi-tab testing was only supported in manual test creation.
Coming soon
- Mailbox steps - Test email-driven workflows, including verifying email delivery, extracting message content, and clicking links.
- JavaScript snippet steps - Author inline JavaScript steps for custom logic, assertions, and data manipulation.
- Parallel agentic test authoring (Labs) - Generate multiple tests simultaneously in mabl’s cloud infrastructure.
Learn more
To learn more, see our guide on the agentic test authoring for web apps.