Failure clustering

The practice of grouping automated test failures by their likely cause after a test run, in order to identify patterns rather than treating each failure in isolation. Common clusters include flakiness (timing issues, async waits), environment problems (service outages, config drift), test data issues (missing accounts, expired tokens), selector failures (DOM changes), and assertion problems (checks that are too broad or brittle). AI tools can propose these clusters automatically, but a human tester should verify the groupings before acting on them.