review

Reviews code changes for naming quality, UI accessibility, and security vulnerabilities.

Updated Jan 30, 2026
One-click install
npx skills add https://github.com/RyoMa99/chezmoi_dotfiles --skill review-ryoma99
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: review
Source: https://github.com/RyoMa99/chezmoi_dotfiles/tree/main/dot_claude/skills/review
Command: npx skills add https://github.com/RyoMa99/chezmoi_dotfiles --skill review-ryoma99

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Code reviews often miss subtle issues across naming design, accessibility, and security. This Skill automates a structured three-perspective review of changed files so problems are caught before a pull request is created. ## Core Features & Use Cases - Naming Review: Evaluates identifiers against a 7-stage naming maturity model (Missing through Domain Abstraction), detecting vague suffixes like Manager/Handler, primitive obsession, and Feature Envy. - UI Check: Runs only when UI component files change, verifying aria attributes, keyboard support, 8pt grid spacing, contrast ratios, and 48px touch targets against the Vercel Web Interface Guidelines. - Security Review: Detects hardcoded secrets, missing input validation, SQL/OS command injection, XSS, authentication gaps, error information leakage, and runs pnpm audit on new dependencies. - Use Case: Before opening a PR, run the review on your git diff to receive severity-ranked findings (Critical/Major/Minor) with concrete fix suggestions and a summary table. ## Quick Start Ask the AI to review the current uncommitted changes for naming, UI, and security issues using the review skill.

Frequently Asked Questions about review

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

FAQPage Schema
How do I review code changes before creating a pull request?▼

Run the review skill against your git diff to analyze changed files across naming, UI accessibility, and security. It detects changed files automatically via git diff when no argument is given, then outputs severity-ranked findings with fix suggestions.

What security vulnerabilities does automated code review detect?▼

It detects hardcoded secrets, missing input validation, SQL and OS command injection, XSS via unsafe HTML insertion, missing authentication middleware, error stack leakage, and vulnerable dependencies via pnpm audit. Findings are classified as Critical, Major, or Minor.

Does the UI accessibility check run on every review?▼

No, the UI check only runs when changed files include UI components such as .tsx, .jsx, .vue, or .svelte files. If no UI files changed, the step is skipped and the skip is explicitly noted in the summary.

What is the 7-stage naming model used in code review?▼

It is a naming maturity ladder from Missing and Nonsense through Honest, Honest and Complete, Does the Right Thing, Intent, and Domain Abstraction. Reviewers classify each identifier and propose incremental improvements rather than jumping to a final name.

Can the review skill modify my code automatically?▼

No, the skill is review-only and never edits code. It outputs problems, reasons, and suggested fixes, and when Critical security findings exist it blocks subsequent implementation steps until you decide how to proceed.