Launch HN: Momentic (YC W24) – Agents that catch bugs by using your app

Mo is an AI-powered Quality Assurance (QA) tool that tests your application through end-to-end flows. It allows users to verify critical functionalities by providing prompts. Trusted by production teams for continuous shipping, Momentic offers a seamless developer experience.

The setup process is straightforward, running tests in parallel 6.3 times faster than traditional methods. This scalability enables Momentic to run tests for every pull request. The result is an enhanced developer experience, contributing to better sleep quality for the team.

By integrating Momentic into the development process, it accelerates the writing and maintenance of end-to-end tests, making the software development cycle more efficient. A case in point is a startup that experienced a 3x reduction in the time and effort required to create and maintain tests.

An automated evaluation of 110 prompts, which would have taken over an hour manually, was completed in just 14 minutes, demonstrating the tool's efficiency. Users praise Momentic for making code move faster, enabling verification to be integrated into the agent loop, and allowing for the reuse of successful AI actions to propose repairs.

Momentic's approach involves writing user flows in plain English within YAML files, which are then reviewed and versioned in the repository. The tool can inspect the application undergoing changes, and when the UI updates, Momentic identifies and proposes fixes, ensuring permanent healing of the issues.

The tool's knowledge base allows teams to describe their product's terminology, business rules, and expected behavior. This context helps the agent choose appropriate actions and verify outcomes. Users have reported catching bugs that might have gone unnoticed in manual testing, citing examples of their products being validated even when deterministic AI responses were not guaranteed.