semantic-commit

Splitting large changes into meaningful commits involves detecting logical boundaries across files and grouping related modifications by feature. This Skill automatically segments multi-file changes, classifies them by type (feature, fix, docs, chore), and generates Conventional Commits messages that reflect the intent of each grouped change, creating a clearer commit history.

Conventional Commits is a standardized format for commit messages that enables automated changelog generation, semantic versioning, and clearer project history. It uses a structure like `type(scope): description` to classify changes. Using it across your team ensures consistency, makes history searchable, and improves code review clarity.

Yes, this Skill is designed for multi-file, multi-feature codebases. It detects change size thresholds (e.g., 5+ files or 100+ lines changed), groups modifications by feature boundaries, and generates focused commits for each logical unit, making it practical for large projects with complex change patterns.

Automatic change segmentation analyzes modified files and line-level changes to identify logical groupings within a large changeset. It uses threshold-based detection and contextual analysis to recognize when modifications belong to the same feature, fix, or documentation effort, then creates separate commits accordingly.

Yes, this Skill detects existing conventions in your repository and tailors commit message generation to match them. This ensures new commits align with your team's established style and standards rather than imposing a generic format.

Before running this Skill, ensure your Git repository contains staged or uncommitted changes you want to segment. The Skill analyzes your latest changes to detect logical boundaries and generate appropriate commit messages, so having a clear, discrete set of modifications produces the best results.