code-review

Review code changes for correctness and edge cases with verification evidence.

Updated Mar 7, 2026
One-click install
npx skills add https://github.com/MinhHoangDono/antigravity-kit --skill code-review-minhhoangdono
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: code-review
Source: https://github.com/MinhHoangDono/antigravity-kit/tree/main/.agent/skills/code-review
Command: npx skills add https://github.com/MinhHoangDono/antigravity-kit --skill code-review-minhhoangdono

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

It helps you perform technically rigorous code reviews that catch edge cases and prevent incorrect or unverified claims from slipping into PRs and releases.

Core Features & Use Cases

  • Evidence-based review workflow: Enforces verification before any completion or status claim.
  • Edge case scouting: Guides reviewers to actively search for boundary failures after changes land.
  • PR/changes focused execution: Supports reviewing recent changes by default or scanning the entire codebase when broader risk exists.
  • Use Case: After implementing a feature, run this Skill to review the diff, validate critical behavior with fresh evidence, and request targeted fixes before merge.

Quick Start

Use the code-review skill to review your most recent changes before claiming the task is complete.

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 rigorous code review on a pull request?

Perform a rigorous code review by validating critical behavior with fresh evidence, scouting for boundary failures in the diff, and requesting targeted fixes based on YAGNI, KISS, and DRY principles.

What is evidence-based code review?

Evidence-based code review enforces verification before any completion claim by running explicit verification commands and actively searching for boundary failures to ensure correctness.

How do I find edge cases in a recent code change?

Find edge cases in a recent code change by applying a verification-first workflow that actively scouts for boundary failures and validates critical behavior with fresh evidence before merge.

When should I scan the entire codebase versus just the diff during a code review?

Review recent changes by default for focused execution, but scan the entire codebase when broader risk exists, ensuring edge cases and defects are caught based on the change scope.

How do I validate a feature implementation before claiming completion?

Validate a feature implementation by running a post-implementation validation that reviews the diff, validates critical behavior with fresh evidence, and requests targeted fixes before merge.

What principles should guide pull request reviews to prevent defects?

Guide pull request reviews with YAGNI, KISS, and DRY principles, emphasizing red-flag detection and explicit verification commands to ensure only verified changes are merged.