How do I install the Code Review skill?

Run npx skillmds@latest add hi-donwi/code-review in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

Is Code Review safe to use?

This skill has not completed SkillMD's automated safety review yet. Capability flags: docs only. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Code Review?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is Code Review free to use?

Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.

Who published Code Review?

hi-donwi (@hi-donwi) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Code Review by hi-donwi — Agent Skill

Name: Code Review
Rating: 4.5 (4 reviews)
Author: hi-donwi

Code Review

Review a diff or PR across correctness, design, tests, security, and clarity, and give actionable, prioritized feedback. Use when asked to review code, before merging a change, or to self-review a diff. Do not use to author large new features from scratch (use the build skills) — this reviews existing changes.

hi-donwi Updated Sep 22, 2026

File contents

Code Review & Quality

Overview

Evaluate a change against multiple axes and return specific, prioritized, actionable feedback.

When to use

Reviewing a PR or working-tree diff before merge.
Self-reviewing your own change before pushing.

Process

Understand intent — read the spec/PR description; review the diff against its stated goal.
Review five axes:
- Correctness — logic, edge cases, error handling, concurrency.
- Design — fits existing patterns; right abstraction; no needless coupling.
- Tests — meaningful coverage for the change and its edge cases.
- Security — input validation, authz, secrets, injection (defer deep checks to security-hardening).
- Clarity — names, readability, comments where non-obvious; matches surrounding style.
Prioritize findings: blocking → should-fix → nit. Be specific and cite file:line.
Size check — flag changes too large to review well; suggest splitting.
Prefer concrete suggestions over vague critique.

Architecture review (reusability standards)

The workspace defines four classifications for reusable code (see AGENTS.md → Architecture Conventions). During review, flag violations:

Violation	Should Be
Business logic inside a UI component	Extracted to a Service or Hook
Duplicated state/effect logic across components	Extracted to a Custom Hook
Inline API calls in page or component files	Delegated to a Service
Framework-specific code mixed with pure logic	Separated into a framework-independent Utility
Reusable workflow embedded in a page file	Moved to a Service or Hook

Red flags (in the change)

New dependencies without justification; reinventing existing utilities.
Untested behavior changes; broad refactors mixed with logic changes.
Swallowed errors, any/weak typing, copy-paste duplication.
Architecture violations from the reusability standards checklist above.

Verification

Every blocking finding cites a location and a concrete fix.
Feedback is prioritized; nits are labeled as nits.

Reference Index

references/expanded-guidance.md — deeper review rubric, structural remedies, sizing guidance, and quality gates.

---name: code-reviewdescription: Review a diff or PR across correctness, design, tests, security, and clarity, and give actionable, prioritized feedback. Use when asked to review code, before merging a change, or to self-review a diff. Do not use to author large new features from scratch (use the build skills) — this reviews existing changes.---# Code Review & Quality## OverviewEvaluate a change against multiple axes and return specific, prioritized, actionable feedback.## When to use- Reviewing a PR or working-tree diff before merge.- Self-reviewing your own change before pushing.## Process1. **Understand intent** — read the spec/PR description; review the diff against its stated goal.2. **Review five axes:** - **Correctness** — logic, edge cases, error handling, concurrency. - **Design** — fits existing patterns; right abstraction; no needless coupling. - **Tests** — meaningful coverage for the change and its edge cases. - **Security** — input validation, authz, secrets, injection (defer deep checks to `security-hardening`). - **Clarity** — names, readability, comments where non-obvious; matches surrounding style.3. **Prioritize** findings: blocking → should-fix → nit. Be specific and cite `file:line`.4. **Size check** — flag changes too large to review well; suggest splitting.5. **Prefer concrete suggestions** over vague critique.## Architecture review (reusability standards)The workspace defines four classifications for reusable code (see `AGENTS.md` →_Architecture Conventions_). During review, flag violations:| Violation | Should Be ||---|---|| Business logic inside a UI component | Extracted to a Service or Hook || Duplicated state/effect logic across components | Extracted to a Custom Hook || Inline API calls in page or component files | Delegated to a Service || Framework-specific code mixed with pure logic | Separated into a framework-independent Utility || Reusable workflow embedded in a page file | Moved to a Service or Hook |## Red flags (in the change)- New dependencies without justification; reinventing existing utilities.- Untested behavior changes; broad refactors mixed with logic changes.- Swallowed errors, `any`/weak typing, copy-paste duplication.- Architecture violations from the reusability standards checklist above.## Verification- Every blocking finding cites a location and a concrete fix.- Feedback is prioritized; nits are labeled as nits.## Reference Index- `references/expanded-guidance.md` — deeper review rubric, structural remedies, sizing guidance, and quality gates.

Frequently asked questions