investigate

To find a bug's root cause, reproduce the issue consistently and trace data flow from user action through event handlers, services, APIs, and databases to identify where data first diverges. Form explicit hypotheses and validate them with logs or a debugger before changing code.

Tracing data flow across frontend and backend systems involves tracking client-to-server-to-database paths. You reproduce the issue, monitor logs or console outputs at each layer, and verify where the expected behavior first diverges from the actual data returned.

Hypothesis-driven testing for intermittent failures requires forming explicit assumptions about the root cause. You validate each hypothesis using logs, console outputs, or a debugger to confirm the exact failure point before implementing any code changes.

You should escalate a bug investigation after three failed attempts to fix the issue. This three-strike policy stops repeated failed fixes, ensuring you halt attempts and escalate the problem when your hypotheses are not resolving the root cause.

To add a regression test during bug triage, write a failing test that specifically captures the reported defect. You verify the root cause with logging, implement the smallest possible fix, and ensure the previously failing test now passes.