test-reliability

Fix flaky Playwright tests by classifying root causes like timing shifts or broken locators, then applying selector healing, data regeneration, and environment-aware diagnosis to produce evidence-backed repairs instead of unreliable blind retries.

Flaky end-to-end tests fail intermittently due to timing variations, expired test data, broken locators, test order dependencies, visual rendering shifts, or inconsistent CI environments requiring environment-aware diagnosis to isolate.

Heal broken Playwright selectors at runtime using multi-attribute selector repair, which updates locators dynamically and validates intent-fidelity to ensure the test targets the correct application element.

Manage quarantined tests in a CI pipeline by routing unstable tests to a quarantine list with confidence scores and evidence logging, ensuring fixes are reviewable while preventing false confidence from masked failures.

Playwright tests pass locally but fail in CI due to environment inconsistencies, timing variations, or expired test data, requiring environment-aware diagnosis and data regeneration to stabilize execution across differing environments.