What problem does it solve?
It solves the problem of turning an active coding session into a clean, accurate git commit by inspecting what actually changed, choosing a conventional commit message, and committing safely.
Core Features & Use Cases
- Session-aware review: checks the current conversation and any available session history to understand intent before drafting the commit message.
- Repository-state inspection: verifies the real changes using
git status --short and relevant diffs so the message matches the code.
- Conventional commit message generation: produces an imperative, conventional commit header with appropriate type/scope and optional body/footers when needed.
- Safe execution workflow: stages everything with
git add -A and creates a commit unless a real blocker or no-changes condition applies; reports the resulting commit header and summary.
Quick Start
Ask the AI to commit the current work with a best-practice Conventional Commits message after reviewing the session context and the repo diffs.