architect-diagnose-test-flakes

To diagnose flaky tests causing intermittent CI failures, you analyze CI logs, recent git commits, and runtime context to classify root causes like timing, test order, shared state, or external dependencies, then build a deterministic reproduction.

Test flakiness and unpredictable automated test failures are caused by timing issues, test execution order, shared mutable state, or external dependency instability, which require evidence-driven categorization to identify and resolve correctly.

Fix flaky tests by constructing a minimal deterministic reproduction command, applying the smallest reliable code change, and adding a guard test with risk and rollback guidance to prevent future regressions.

Yes, you can use CI logs and git history to troubleshoot unstable test pipelines by turning failing logs, recent code changes, and runtime context into a structured investigation that identifies the exact flake root cause.

The best way to stop repeated test reruns and stabilize CI behavior is performing root cause analysis on intermittent failures to distinguish timing, order, and state issues, then implementing a concrete fix and guard test plan.

Investigate test order and shared state for CI instability when the same test fails unpredictably, retries are frequently needed, or timing and external dependencies are suspected as the source of flakiness.