Finalrun - Spec-driven testing for mobile apps using English and vision
Finalrun is an open-source tool that enables testing mobile apps using plain English and vision-based execution, keeping tests in sync with the codebase.
Why it matters
Finalrun's approach to mobile app testing can improve reliability and maintainability by keeping tests in sync with the codebase.
Key Points
- 1Vision-based agent can understand intent and perform actions across Android and iOS
- 2Tests are generated directly from the codebase to stay in sync and have more context
- 3YAML-based test flows and vision-based execution without brittle selectors
- 4Demonstration of AI-built feature with immediately generated and executed vision-based test
Details
Finalrun is a tool that aims to simplify mobile app testing by allowing tests to be defined in plain English instead of relying on fragile selectors like XPath or accessibility IDs. The core idea is to use a vision-based agent that can understand the intent of the test and perform actions directly on the app's user interface, across both Android and iOS platforms. The key innovation is in how the tests are defined and maintained. Rather than keeping the test flows outside the codebase, which can lead to them going out of sync over time, Finalrun generates the tests directly from the codebase context. This helps ensure the tests stay up-to-date and have more relevant information. The tests are defined in a YAML-based format, which provides a readable and maintainable approach. The demonstration video shows how Finalrun can be used in a 'post-development hand-off' scenario, where an AI-built feature is immediately tested using a vision-based test generated by the tool. This highlights the potential for tighter integration between AI-powered development and automated testing.
No comments yet
Be the first to comment