New to Claude Skills? Learn how to install them β†’

juliusbrussee on GitHub

Caveman Review

Free

Streamlined code review comments for clarity and action.

Get this skill

Free Β· Opens the source repo

What Caveman Review does

Caveman Review is designed to enhance the code review process by providing ultra-compressed feedback that prioritizes actionable insights. Each comment generated by this skill follows a strict one-line format, ensuring that reviewers can quickly identify issues without sifting through unnecessary commentary. The format includes the line number, the identified problem, and a suggested fix, making it easy for developers to understand and address issues in their pull requests. This skill is particularly useful for teams that want to maintain high code quality while minimizing the time spent on reviews.

The skill operates under a set of clear rules that dictate how comments should be structured. It emphasizes brevity and clarity by avoiding common phrases that add little value, such as hedging or unnecessary praise. Instead, it focuses on delivering precise feedback that can be directly acted upon. For example, instead of a verbose explanation of a potential bug, a comment might simply state the line number, the severity of the issue, and a concise fix. This approach not only saves time but also helps maintain the focus on the code itself.

Caveman Review is ideal for development teams looking to streamline their code review process. It is especially beneficial in environments where quick iterations and rapid feedback are essential, such as agile development. By using this skill, teams can ensure that their code reviews are efficient and that important issues are highlighted without the clutter of unnecessary commentary.

However, it’s important to note that Caveman Review is strictly limited to providing comments and does not offer code fixes or approval requests. It is meant solely for reviewing code and generating comments that can be pasted directly into pull requests. This makes it less suitable for teams that require a more comprehensive review process that includes discussions or deeper analyses of architectural decisions.

When to use it

Use this skill when you need concise and clear feedback on pull requests, especially in fast-paced development environments.

When not to use it

This skill is not suitable for situations requiring detailed explanations or discussions about architectural decisions.

What you can build with it

Rapid PR Feedback

Use Caveman Review to quickly generate actionable comments on pull requests, saving time during code reviews.

Focus on Actionable Insights

Leverage the skill to eliminate unnecessary commentary and focus on what needs to be fixed in the code.

Streamlined Review Process

Integrate Caveman Review into your development workflow to enhance efficiency and maintain high code quality.

How to install Caveman Review

View source

1. Install with the skills CLI

npx skills add juliusbrussee/caveman/caveman-review --agent claude-code

2. Or install it manually

Download the skill folder and drop it into ~/.claude/skills/ for all projects, or .claude/skills/ to scope it to one repo. Restart Claude Code so it picks up the new skill.

Anthropic's agentic coding CLI, and the reference implementation of Agent Skills. Drop a skill folder into ~/.claude/skills and Claude Code loads it automatically whenever a task matches the skill's description. Claude Code docs

Inside SKILL.md

Written by juliusbrussee

Write code review comments terse and actionable. One line per finding. Location, problem, fix. No throat-clearing.

Rules

Format: L<line>: <problem>. <fix>. β€” or <file>:L<line>: ... when reviewing multi-file diffs.

Severity prefix (optional, when mixed):

  • πŸ”΄ bug: β€” broken behavior, will cause incident
  • 🟑 risk: β€” works but fragile (race, missing null check, swallowed error)
  • πŸ”΅ nit: β€” style, naming, micro-optim. Author can ignore
  • ❓ q: β€” genuine question, not a suggestion

Drop:

  • "I noticed that...", "It seems like...", "You might want to consider..."
  • "This is just a suggestion but..." β€” use nit: instead
  • "Great work!", "Looks good overall but..." β€” say it once at the top, not per comment
  • Restating what the line does β€” the reviewer can read the diff
  • Hedging ("perhaps", "maybe", "I think") β€” if unsure use q:

Keep:

  • Exact line numbers
  • Exact symbol/function/variable names in backticks
  • Concrete fix, not "consider refactoring this"
  • The why if the fix isn't obvious from the problem statement

Examples

❌ "I noticed that on line 42 you're not checking if the user object is null before accessing the email property. This could potentially cause a crash if the user is not found in the database. You might want to add a null check here."

βœ… L42: πŸ”΄ bug: user can be null after .find(). Add guard before .email.

❌ "It looks like this function is doing a lot of things and might benefit from being broken up into smaller functions for readability."

βœ… L88-140: πŸ”΅ nit: 50-line fn does 4 things. Extract validate/normalize/persist.

❌ "Have you considered what happens if the API returns a 429? I think we should probably handle that case."

βœ… L23: 🟑 risk: no retry on 429. Wrap in withBackoff(3).

Auto-Clarity

Drop terse mode for: security findings (CVE-class bugs need full explanation + reference), architectural disagreements (need rationale, not just a one-liner), and onboarding contexts where the author is new and needs the "why". In those cases write a normal paragraph, then resume terse for the rest.

Boundaries

Reviews only β€” does not write the code fix, does not approve/request-changes, does not run linters. Output the comment(s) ready to paste into the PR. "stop caveman-review" or "normal mode": revert to verbose review style.

Frequently asked questions about Caveman Review

Similar skills