code-review

Review code changes across correctness, security, performance, and maintainability layers.

2|Updated Apr 30, 2026

Search Tags:#security #maintainability #static-analysis #quality-assurance #code-review #pull-request

Ready to run

npx skills add https://github.com/BOHUYESHAN-APB/openagent-labforge-bio --skill code-review-bohuyeshan-apb

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: code-review
Source: https://github.com/BOHUYESHAN-APB/openagent-labforge-bio/tree/main/src/skills/code-review
Command: npx skills add https://github.com/BOHUYESHAN-APB/openagent-labforge-bio --skill code-review-bohuyeshan-apb

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Systematic code review across 4 layers with severity classification to surface meaningful issues while filtering low-confidence findings.

Core Features & Use Cases

Four-layer analysis: Correctness, Security, Performance, and Style & Maintainability to catch a wide range of concerns.
Severity-driven actions: Critical/Major/Minor/Nitpick risk levels with recommended responses and prioritization.
Defined workflow: Stepwise review process with initial scan, deep analysis, context checks, and synthesized reporting.

Quick Start

Provide the code changes or a PR link to run a layered review and receive confidence-filtered findings.

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 for a pull request?▼

To perform a structured code review, provide your pull request link or code changes to run a multi-layer analysis covering correctness, security, performance, and maintainability, returning confidence-filtered findings with file:line references.

What is multi-layer static analysis for code quality assurance?▼

Multi-layer static analysis for code quality assurance systematically evaluates source code across four dimensions—correctness, security, performance, and style—to surface meaningful issues while filtering out low-confidence findings.

How does severity classification work in automated code reviews?▼

Severity classification in automated code reviews categorizes identified issues into Critical, Major, Minor, and Nitpick risk levels, providing recommended responses and prioritized actions to help developers address the most impactful problems first.

Does this code review process enforce a confidence threshold for findings?▼

Yes, this code review process enforces a strict 80% confidence threshold, ensuring that only high-certainty issues regarding correctness, security, performance, and maintainability are reported to filter out low-confidence findings.

Can I use this for auditing existing codebases instead of just new changes?▼

Yes, you can use this for auditing existing codebases, as it applies to new changes, pull requests, and comprehensive audits where robust correctness, security, performance, and maintainability checks are needed.

What is the defined workflow for a stepwise code review?▼

The defined workflow for a stepwise code review involves an initial scan, deep analysis, context checks, and synthesized reporting to systematically identify and report code quality issues across multiple layers.