TestTrace

TestTrace is a web app (with a lightweight CI agent) that turns “flaky test” arguments into actionable, evidence-backed root-cause reports. It ingests CI artifacts—logs, screenshots, videos, network traces, and timing metrics—then correlates failures across runs, branches, and environments to pinpoint what changed and what’s unstable. Instead of generic “rerun it” advice, it produces a timeline of events, highlights nondeterministic signals (timeouts, race conditions, async waits), and flags environment drift (browser versions, CPU throttling, dependency updates). It also auto-tags failures into buckets like “selector instability,” “backend latency,” “test data collisions,” and “parallelism interference,” so teams can prioritize fixes. This is a combination traditional + AI app: traditional pipelines for ingestion/correlation and AI to summarize evidence and suggest the most likely failure mechanism. It’s not magic—if you don’t collect artifacts, it can’t help.