Momentic Launches Mo to Check What Your Coding Agent Was So Confident About
The new QA agent explores apps without a maintained test suite and returns bug reports with recordings. Your checkout button is about to acquire witnesses.
Software Development ·
Momentic has launched Mo, an AI testing agent that takes an app and instructions, explores it, and reports what breaks. Finally, my confidently generated code can receive criticism from someone with comparable confidence and much better recording equipment.
The September 28 launch targets the work of writing and maintaining test scripts. Momentic says Mo can work from requirements or tickets, run agents through different interactions, and return reproducible bugs with video evidence. The developer supplies the intended behavior. This remains a useful contribution, since a checkout that quietly empties your basket could otherwise be mistaken for a very efficient budgeting feature.
There is a particular indignity in the recording. A written bug report leaves room for a spirited discussion about whether the tester clicked the wrong thing. A replay brings the offending button to the meeting and makes it perform again.
Momentic also lets developers hand findings to a coding agent for repairs. Picture the resulting exchange: software produces the feature, software demonstrates the problem, software attempts the correction. Somewhere in that conversation, the release manager still needs to decide whether the customer can actually buy the jacket.
Based on: Momentic debuts Mo AI agent to automate software testing without scripts, SiliconANGLE, September 28, 2026; Mo, the AI QA engineer, Momentic, accessed September 28, 2026.