kata-stacked-pr-rebase

To rebase stacked PRs, you must propagate changes from the modified lower layer up through all upper layers in a bottom-to-top order, rebasing each branch on the updated previous layer and using lease-based force pushes to maintain linear history.

The correct cascade order for stacked pull requests is strictly bottom-to-top, ensuring each upper layer is rebased on top of the updated previous layer to preserve intended history without skipping intermediate changes.

During a stacked PR rebase cascade, conflict resolution requires manual intervention to resolve discrepancies, followed by continuation or abort commands, and final verification checks to ensure the history matches expectations.

Yes, stacked PR rebase addresses the special case of upstream squash-merged divergence by applying cascading updates to multi-layer feature branch stacks in git worktrees while preserving the intended linear history.

You use force-with-lease when pushing rebased stacked PR branches to safely update remote references and avoid overwriting others' work, providing a guardrail against accidental data loss during the cascade rebase process.

When cascading rebase across git worktree branches, you should avoid rebasing main within the cascade flow and ensure final verification checks are performed to confirm the resulting history matches your expectations.