visual-verdict

Automated screenshot comparison for UI regression testing is achieved by passing your reference images and a generated UI screenshot to generate a deterministic JSON verdict. This verdict includes a visual score, differences, and reasoning to provide repeatable QA feedback.

A deterministic visual verdict is a strict JSON report identifying visual differences between a UI screenshot and reference images. It outputs consistent fields like score, category match, and suggestions to enable fast, repeatable visual acceptance decisions.

Yes, you can supply multiple reference images to evaluate against a single generated screenshot. The workflow processes these inputs to output a JSON verdict, helping you verify if a design iteration passes visual checks across various expected conditions.

You integrate visual diff checks into a CI pipeline by running the screenshot comparison workflow during automated builds. It evaluates the generated screenshot against reference images and returns a JSON verdict to quickly decide if the UI passes visual acceptance.

No external dependencies are required to run image diff evaluation. The workflow operates independently to process your reference images and generated screenshots, directly producing a JSON verdict with score, differences, and suggestions.

The visual verdict outputs a structured JSON format for automated QA. This JSON includes specific fields such as score, verdict, category match, differences, suggestions, and reasoning to facilitate strict visual difference reporting.