skills.rest
navikt
naviktOFFICIAL
Explore all skills by this author →
WebsiteGitHub

caveman-review

Generate one-line code-review comments with severity prefixes for diffs.

1|1|Updated Apr 9, 2024
Search Tags:#automation#commenting#formatting#diff#code-review#pull-request#line-numbers
One-click install
npx skills add https://github.com/navikt/helved-utbetaling --skill caveman-review-navikt
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: caveman-review
Source: https://github.com/navikt/helved-utbetaling/tree/main/.agents/skills/caveman-review
Command: npx skills add https://github.com/navikt/helved-utbetaling --skill caveman-review-navikt

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Ultra-compact code review comments that cut noise from PR feedback while preserving actionable signal. Each comment is one line: location, problem, fix. Use when user says "review this PR", "code review", "review the diff", or invokes /caveman-review. Auto-triggers when reviewing pull requests.

Core Features & Use Cases

  • One-line-per-finding comments with exact line numbers and location references.
  • Severity prefixes to categorize issues (🔴 bug:, 🟡 risk:, 🔵 nit:, ❓ q:).
  • Drop/keep rules to reduce boilerplate and keep the signal clear.
  • Works across multi-file diffs and enforces exact identifiers.

Quick Start

Review a PR diff and return one-line, actionable comments for each finding in the format 'L<line>: <problem>. <fix>.'

Frequently Asked Questions about caveman-review

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I generate compact code review comments for a pull request diff?▼

Generate compact code review comments by outputting one-line findings with exact line numbers, severity prefixes, and drop/keep rules. Each comment follows the format 'L<line>: <problem>. <fix>.' to ensure actionable feedback across multi-file diffs.

How does formatting code review feedback with severity prefixes improve pull request reviews?▼

Severity prefixes categorize issues into bugs, risks, nits, and questions, reducing boilerplate through drop/keep rules. This enforces exact line references and code identifiers across multi-file diffs, keeping the signal clear and actionable.

Can I use automated code review comments across multi-file diffs in different programming languages?▼

Yes, automated code review comments apply to multi-file diffs across various programming languages. The system enforces exact line references and code identifiers, handling scenarios like style fixes, bug detection, and refactoring notes.

What is the best way to automate terse code review comments for bug detection and style fixes?▼

The best way to automate terse code review comments is to enforce one-line-per-finding outputs with exact line numbers and severity prefixes. Drop/keep rules reduce boilerplate, ensuring feedback remains actionable for style fixes, bug detection, and refactoring.

What are the limitations of using one-line code review comments for complex pull request reviews?▼

One-line code review comments are limited to terse, actionable feedback, enforcing exact line references and severity prefixes. They may not capture nuanced architectural discussions or deep refactoring strategies that require longer explanations across multi-file diffs.

Related Skills

Explore More Software Engineering Skills→
hairihouhairihou

conventional-comments

Generate code review comments in the Conventional Comments format.

Community
Intermediate
richardadonnellrichardadonnell
45

sentry-pr-code-review

Parse Seer by Sentry bot comments on GitHub pull requests to extract bugs and suggested fixes.

Community
Intermediate
wiltaylorwiltaylor

review-pr

Generate structured JSON feedback from pull request diffs.

Community
Intermediate
giovannimnzgiovannimnz

caveman-review

Generate one-line code-review comments for PR diffs with severity prefixes and line numbers.

Community
Intermediate