Cursor Code Review Assistant

Review AI-generated code for quality and suggest improvements to your AI rules based on output quality. Ensure code meets team standards and continuously improve AI-assistant configurations.

Validate AI-generated code against team standards before committing. The skill analyzes code quality, identifies patterns in issues, suggests improvements to your Cursor AI rules, and tracks quality metrics over time to drive continuous improvement of AI-generated code.

code-reviewcursorqualitystandardsai-rules

Version 1.0.0Updated Jan 27, 2026

Capabilities

Pre-commit quality-review process

Team-standards validation

AI rules improvement suggestions

Quality-metrics tracking

Context-aware code analysis

Use cases

Reviewing generated code before commit

Checking against team coding standards

Improving Cursor AI rules configuration

Tracking code-quality trends

Setting up automated code-review workflows

Try it

Example prompts to use with this skill

Prompt

Review the authentication code I just generated before I commit

Output

## Code Review: Authentication Module ### ✅ Strengths - Proper TypeScript types used throughout - Password hashing implemented correctly - JWT token generation follows best practices ### ⚠️ Issues Found 1. Missing error handling in `login()` function - Line 45: API call not wrapped in try-catch - Recommendation: Add error handling with specific error types 2. Hard-coded JWT secret - Line 23: Secret stored in code - Recommendation: Move to environment variables ### 📋 Before Committing - [ ] Fix error handling in login function - [ ] Move JWT secret to .env - [ ] Add rate limiting (or create follow-up ticket) - [ ] Add tests for error cases

Prompt

Check if the new API endpoints follow our team standards

Output

✅ PASS: TypeScript strict mode ✅ PASS: Async/await used consistently ✅ PASS: Proper error handling ⚠️ WARN: Some files with more than 200 lines (api/users.ts: 245 lines) ❌ FAIL: Missing input validation on three endpoints ❌ FAIL: No authentication checks on public routes ## Required Changes: 1. Add input validation using Zod or Joi 2. Split large files into smaller modules 3. Add authentication middleware to protected routes

Add to your AI assistant

Choose your AI assistant and run the command in your terminal

Cursor

curl -fsSL https://raw.githubusercontent.com/n3wth/newth-skills/main/skills/cursor-code-review.md -o ~/.cursor/skills/cursor-code-review.md

Copy

Claude

curl -fsSL https://raw.githubusercontent.com/n3wth/newth-skills/main/skills/cursor-code-review.md -o ~/.claude/skills/cursor-code-review.md

Copy

Windsurf

curl -fsSL https://raw.githubusercontent.com/n3wth/newth-skills/main/skills/cursor-code-review.md -o ~/.windsurf/skills/cursor-code-review.md

Copy

Discussions

Related skills

New

Code Reviewer

Automated code review with best practices. Identify code smells, suggest refactoring improvements, check for security issues, and enforce coding standards.

code-reviewqualitybest-practices

New

Smart .cursor/rules Generator

Auto-generate AI rules from coding patterns and project characteristics. Create project-specific AI rules that match your team's coding style.

cursorai-rulescode-analysis