code-review

Review code across four layers and report issues with file:line evidence.

Updated Mar 18, 2026
One-click install
npx skills add https://github.com/dmooney/Rundale --skill code-review-dmooney
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: code-review
Source: https://github.com/dmooney/Rundale/tree/main/.opencode/skills/code-review
Command: npx skills add https://github.com/dmooney/Rundale --skill code-review-dmooney

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill helps you perform consistent, noise-controlled code reviews that identify correctness, security, performance, and maintainability issues while only escalating findings you’re highly confident about.

Core Features & Use Cases

  • 4-layer systematic review: evaluates Correctness, Security, Performance, and Style & Maintainability to ensure comprehensive coverage.
  • Severity classification + required action: categorizes findings as 🔴 Critical, 🟠 Major, 🟡 Minor, or 🟢 Nitpick to standardize prioritization and next steps.
  • 80% confidence threshold: only reports issues at high confidence and explicitly labels uncertainty to reduce false positives.
  • Structured output format: produces a consistent review report including files reviewed, overall assessment, grouped findings with file:line references, and positive observations.

Quick Start

Ask the code-review skill to review the latest changes and return a structured report with file:line references for any findings at 80% confidence or higher.

Frequently Asked Questions about code-review

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

FAQPage Schema
How do I perform a structured code review that reduces false positives?

Structured code review reduces false positives by applying an 80% confidence threshold and requiring file:line references for all findings. This process evaluates correctness, security, performance, and maintainability while explicitly labeling uncertainty to minimize noise.

What is the best way to classify code review findings by severity?

Classifying code review findings by severity standardizes prioritization by categorizing issues as Critical, Major, Minor, or Nitpick. This severity classification requires corresponding actions and helps teams address correctness, security, performance, and maintainability issues effectively.

How does a multi-layer code review handle security and performance issues?

A multi-layer code review handles security and performance issues by systematically evaluating each domain independently alongside correctness and maintainability. It produces an evidence-backed report with file:line references, escalating only highly confident findings above the 80% threshold.

Can I use an automated code review for post-change auditing?

Automated code review supports post-change auditing by producing a consistent report that includes files reviewed, an overall assessment, and grouped findings. It applies a 4-layer systematic evaluation to verify implementation completion and identify issues after code changes.

Does code review with a confidence threshold miss low-severity issues?

Code review with an 80% confidence threshold intentionally filters out low-confidence findings to reduce false positives, but it still reports highly confident issues across all severity levels, including Minor and Nitpick categories, ensuring low-severity issues are captured when evidence is strong.