Back to Skills

caveman-commit

Ultra-compressed commit message generator. Cuts noise from commit messages while preserving intent and reasoning. Conventional Commits format. Subject ≤50 chars, body only when "why" isn't obvious. Use when user says "write a commit", "commit message", "generate commit", "/commit", or invokes /caveman-commit. Auto-triggers when staging changes.

26,438stars1,230forksUpdated 4/14/2026

Security Assessment

Safe(100/100)
Security Score100/100

About caveman-commit

The caveman-commit skill is an AI-powered tool designed to generate ultra-compressed commit messages for version control systems. It addresses the common problem of verbose, unclear, or inconsistent commit messages by distilling changes into concise, precise statements that still capture the intent and reasoning behind the code modifications. By adhering to the Conventional Commits standard, it ensures that commit messages are structured, predictable, and compatible with automation workflows, while reducing noise and unnecessary filler text that can obscure the purpose of a change.

This skill offers several key features and capabilities. It automatically generates subject lines in the form `<type>(<scope>): <imperative summary>`, supporting common types such as `feat`, `fix`, `refactor`, `perf`, `docs`, `test`, `chore`, `build`, `ci`, `style`, and `revert`. Subject lines are kept short, ideally under 50 characters, and always in the imperative mood. The body of the commit message is included only when necessary to explain non-obvious reasoning, breaking changes, migrations, or linked issues, with proper formatting including 72-character line wraps and bullet points. The skill strictly avoids including personal pronouns, AI attributions, emojis, or redundant information, and automatically triggers when changes are staged. Users can also manually invoke it via commands like "/commit" or "/caveman-commit".

Caveman-commit is particularly useful for developers, code reviewers, and teams aiming to maintain a clean, readable commit history that communicates purpose without clutter. It is ideal for projects that require strict adherence to Conventional Commits for automation pipelines, CI/CD processes, or changelog generation. The skill is suited for scenarios such as writing commits for new features, refactors, bug fixes, breaking API changes, or security updates, providing clarity and consistency across collaborative development environments.

FAQ

How do I invoke caveman-commit manually?

You can trigger caveman-commit by typing commands like 'write a commit', 'commit message', 'generate commit', '/commit', or '/caveman-commit'. It also auto-triggers when changes are staged.

Does caveman-commit execute git commands or stage files?

No, caveman-commit only generates commit messages. It does not run 'git commit', stage files, or amend commits. The output is provided as a code block ready to paste.

When does caveman-commit include a commit message body?

The body is included for non-obvious 'why' explanations, breaking changes, migration notes, linked issues, security fixes, data migrations, or when reverting a prior commit. Otherwise, the subject line alone is sufficient.

What commit message formats are supported?

It follows the Conventional Commits specification, supporting types like 'feat', 'fix', 'refactor', 'perf', 'docs', 'test', 'chore', 'build', 'ci', 'style', and 'revert', with optional scope and imperative mood.

Are there any limitations on commit message content?

Yes, caveman-commit avoids personal pronouns, AI attributions, emojis (unless required by project convention), and redundant file names if the scope already specifies context.

Install caveman-commit

Download and extract the skill files to your .claude/skills/ directory.

Quick Setup:

  1. Copy the skill folder to .claude/skills/
  2. Claude will automatically detect and use the skill