git-bisect

Git bisect methodically narrows down the culprit commit by testing a range of commits between a known good ref and a bad ref. This automated workflow repeatedly checks out commits and runs your test command to isolate the failure.

Automating flaky test debugging with git bisect requires a bad ref, a good ref, and an executable test command. The workflow then automatically checks out intermediate commits and executes the test command to identify the failing commit.

Git bisect supports both automated bisect runs and manual verification workflows. You can manually mark each checked-out commit as good or bad, allowing you to inspect behavior interactively when an automated test command is not available.

Git bisect handles complex repositories and worktrees by guiding environment setup and test execution. It navigates large commit histories efficiently, applying the automated or manual test workflow across the bisected range to locate the failure.

Running a git bisect workflow requires the Git CLI to be available, a clean repository state, and executable test commands to verify failures. You must also provide a known bad ref and a known good ref to establish the search range.

Git bisect requires a clean repository state to successfully check out intermediate commits without conflicts. Uncommitted changes or untracked files will block the checkout process, preventing the workflow from isolating the failing commit.