squash-commits

Squashing git commits involves merging contiguous thematic groups of commits into clean conventional commit messages. This Skill automates that analysis and provides safety guardrails like original HEAD recording and backup tags to prevent accidental data loss.

The best way to clean up git history is by operating on unpushed commits by default, automatically grouping related changes thematically, and generating conventional commit messages. This approach avoids force pushing and keeps your project history clean.

Squashing commits works with pushed git history by enabling an option to include it, which requires a force push to update the remote. By default, the process only operates on unpushed commits to avoid rewriting shared history.

Generating conventional commit messages when squashing is handled automatically during the thematic grouping process. The analysis identifies the core purpose of contiguous commits and produces clean, imperative messages adhering to conventional commit standards.

Recovering your original git history is possible because the process records the original HEAD and creates backup tags before squashing. It also performs post-squash health checks to catch issues early, ensuring you can revert if needed.

Squashing a large number of git commits in batches is supported through an iterative batch processing mode. This allows you to handle large histories in multiple passes, grouping contiguous thematic commits interactively rather than in a single rewrite.