Browser & computer use
A Playwright reporter sends failed-test context to Jev, categorizing failures as product bugs, automation issues, system issues, or cases to investigate.
Only you can see your notes.
Screenshot unavailable. Open the experiment ↗
Jev supplies a first-pass category, not a definitive diagnosis. Each failed attempt, including retries, creates a model call, sending error context and page data through OpenRouter. The application does not explicitly pin a model, and classification accuracy was not independently tested.
For a developer's existing browser checks, use the report to sort failed checks into possible product bugs, test problems, system problems or cases needing investigation. Jev judges the error context through OpenRouter. Start with a nonsensitive failure and compare the suggested category with what investigation actually finds.
Error and page data leave the local test run. Each failed attempt, including a retry, creates another model call, so repeated failures increase cost. Without an API key, classification is skipped. The chart loads an external library over the internet; neither the chart nor the category proves the cause.