Memory

What is Memory

Memory is how qtrl's AI builds and retains knowledge about your application. As the AI explores your app, runs tests, and executes tasks, it learns how things work: what pages exist, how users navigate between them, what forms and interactions are available, and how different features behave.

This accumulated knowledge makes the AI more effective over time. The more it interacts with your application, the better it understands it, and the more useful its test generation and execution become.

How Memory helps

Memory improves the quality of the AI's work in several ways:

• Better test generation: when the AI knows how your application works, it creates more relevant and thorough test cases. It understands which flows are important, what edge cases exist, and how features connect to each other.
• More reliable execution: the AI navigates your application more confidently when it already knows the layout and behavior of pages. This reduces flaky test results caused by the AI not knowing where to find things.
• Smarter decisions: with context about your application's domain and business logic, the AI can make better judgment calls about what constitutes a pass or fail, and what details are worth checking.

How Memory grows

Memory builds up naturally as you use qtrl. Every task the AI performs contributes to its understanding of your application:

• Exploration tasks are the most direct way to build Memory. When you create an exploration task, the AI deliberately navigates through your application to learn about it. This is a great way to get started with a new project or to update the AI's knowledge after significant changes to your app.
• Test execution also adds to Memory. As the AI runs tests, it observes how pages respond, what workflows look like in practice, and how the application behaves under different conditions.
• Test generation tasks contribute as well, since the AI explores parts of your application while figuring out what tests to create.

Getting the most out of Memory

When you first set up a project, consider running an exploration task before generating tests. This gives the AI a baseline understanding of your application, which leads to better quality test generation from the start.

After major application updates (new features, redesigned pages, changed navigation), running another exploration task helps the AI stay current. The AI will update its knowledge with the latest state of your application.

Memory is project-specific. Each project has its own set of knowledge, which means the AI's understanding is always scoped to the application you are working with.