caveman-review

Generate one-line code review comments with severity tags and fixes.

1|Updated Apr 27, 2026
One-click install
npx skills add https://github.com/dwaynimay/IoTProject --skill caveman-review-dwaynimay
Or copy as Structured Prompt for Agentā–¼
Please help me install this Agent Skill.
Skill: caveman-review
Source: https://github.com/dwaynimay/IoTProject/tree/main/.agents/skills/caveman-review
Command: npx skills add https://github.com/dwaynimay/IoTProject --skill caveman-review-dwaynimay

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

caveman-review eliminates noisy, verbose pull request comments by producing ultra-compressed, paste-ready feedback that clearly states the location, the problem, and the concrete fix.

Core Features & Use Cases

  • One-line, location-first review comments: Produces a single line per finding in the format L<line>: <severity> <problem>. <fix>. to keep reviews scannable.
  • Consistency and severity signaling: Uses severity prefixes (šŸ”“ bug, 🟔 risk, šŸ”µ nit, ā“ question) to help authors triage feedback quickly.
  • Noise reduction and better rationale when needed: Drops hedging and restating the diff, while adding more context for security-grade issues, architectural disagreements, and onboarding scenarios.

Quick Start

Invoke caveman-review by asking the assistant to review the diff or using the command /caveman-review, and it will return one-line comments ready to paste into the PR.

Frequently Asked Questions about caveman-review

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

FAQPage Schema
How do I generate concise code review comments for a pull request?ā–¼

To generate concise code review comments, process the pull request diff to output one line per finding. This format maps each finding to an exact line location, tags it by severity, and proposes a concrete fix.

Can I format pull request feedback to include exact line numbers and severity levels?ā–¼

Yes, pull request feedback can be formatted to include exact line numbers and severity levels. The output strictly follows a single-line format per finding, using severity prefixes like bug, risk, nit, or question for quick triage.

What is the best way to automate noise-free diff analysis for code reviews?ā–¼

Automating noise-free diff analysis requires dropping hedging language and restating of the diff. The analysis produces ultra-compressed feedback that clearly states the location, the problem, and the concrete fix without unnecessary text.

How do I add context to security issues in automated code review feedback?ā–¼

To add context to security issues in code review feedback, expand the single-line format with an optional rationale. This provides the necessary architectural and security context for complex findings while keeping other comments scannable.

Do I need to install dependencies to run lintless code review on a diff?ā–¼

No dependencies are required to run lintless code review on a diff. You can trigger the review directly by asking the assistant to review the diff or by invoking the command interface.

Why use a one-line-per-finding format for pull request code reviews?ā–¼

Using a one-line-per-finding format for pull request code reviews eliminates noisy, verbose comments. This ultra-compressed approach keeps reviews scannable and helps authors triage feedback quickly using severity prefixes.