AI Coding Assistant Customization: Cursor Rules, Windsurf Skills & Claude Projects Complete Guide【2026】

Project-level: .cursor/rules or .cursorrules
Global: Cursor Settings > General > Rules for AI

# Project Conventions

## Code Style
- Use TypeScript strict mode
- Prefer functional programming paradigm
- Use PascalCase for components
- Use camelCase for utility functions

## Tech Stack
- Frontend Framework: React 18 + Next.js 14
- State Management: Zustand
- Styling: Tailwind CSS
- Testing: Vitest + Testing Library

## Comment Guidelines
- Don't add unnecessary comments
- Complex logic requires brief explanations
- Use JSDoc for public API documentation

# Advanced Rules

## File Organization
When creating new components, follow this structure:

## Error Handling
- Use custom Error classes
- API calls must include error boundaries
- Wrap async operations in try-catch

## Security Rules
- Never hardcode secrets in code
- Use environment variables for sensitive config
- User input must be validated

## Response Format
- Explain the approach first
- Provide complete runnable code
- Note potential edge cases

name: React Component Generator
description: Generate React components following project conventions
triggers:
  - "create component"
  - "new component"
  - "generate component"

instructions: |
  When user requests creating a React component:
  
  1. Use functional components with TypeScript
  2. Follow this template structure:
  
  ```tsx
  import { FC } from 'react';
  
  interface ${ComponentName}Props {
    // props definition
  }
  
  export const ${ComponentName}: FC<${ComponentName}Props> = (props) => {
    return (
      <div>
        {/* component content */}
      </div>
    );
  };

### Common Skills Templates

**Code Review Skill**:

```yaml
name: Code Review Assistant
description: Provide professional code review suggestions

instructions: |
  When reviewing code, focus on these aspects:
  
  ## Code Quality
  - Are names clearly expressing intent
  - Do functions follow single responsibility
  - Is there duplicate code
  
  ## Performance Considerations
  - Are there unnecessary re-renders
  - Is data structure choice appropriate
  - Are there memory leak risks
  
  ## Security Checks
  - Is input validation complete
  - Are there XSS/injection risks
  - Is sensitive data handled securely
  
  Output format:
  1. Issue description
  2. Risk level (High/Medium/Low)
  3. Fix recommendation
  4. Fixed code example

name: API Endpoint Generator
description: Generate RESTful API endpoints

instructions: |
  When creating API endpoints, follow these conventions:
  
  - Use RESTful naming conventions
  - Include comprehensive error handling
  - Add request parameter validation
  - Generate OpenAPI documentation comments
  - Include unit test examples

# Project Background
This is an enterprise SaaS application backend service using:
- Language: Python 3.11
- Framework: FastAPI
- Database: PostgreSQL + Redis
- Deployment: Kubernetes

# Code Standards
- Follow PEP 8 style guide
- Use type annotations
- Write docstring documentation
- Unit test coverage > 80%

# Response Preferences
- Prioritize code maintainability
- Explain design decision rationale
- Provide multiple implementation comparisons
- Note potential performance impacts

# Role Definition
You are a senior Python backend engineer focusing on:
- High-performance API design
- Database optimization
- System architecture design

# Task Processing Flow
1. Understand requirements, confirm key points
2. Analyze existing code structure
3. Propose solutions
4. Write implementation code
5. Add test cases

# Output Requirements
- Code must be directly runnable
- Include necessary import statements
- Add error handling logic
- Consider edge cases

# AI Coding Assistant Configuration Template

## Project Information
- Project Name: [project name]
- Tech Stack: [framework/language/tools]
- Project Type: [Web app/API service/CLI tool]

## Code Standards
- Naming Convention: [camelCase/snake_case/PascalCase]
- Indentation: [spaces/tabs]
- Max Line Length: [80/100/120]
- Quote Style: [single/double]

## AI Behavior Preferences
- Explanation Detail: [concise/moderate/detailed]
- Code Comments: [none/key areas/detailed]
- Error Handling: [basic/complete/defensive]
- Test Generation: [none/unit tests/full coverage]

## Prohibited Actions
- Don't use deprecated APIs
- Don't hardcode sensitive information
- Don't ignore error handling
- Don't generate overly long functions

# Team Shared Rules

## Git Commit Standards
Commit message format: <type>(<scope>): <description>
- feat: New feature
- fix: Bug fix
- docs: Documentation update
- refactor: Code refactoring

## PR Description Template
When generating PRs, include:
1. Change summary
2. Testing notes
3. Screenshots (if applicable)
4. Related Issues

## Code Review Checklist
- [ ] Code follows project standards
- [ ] Includes necessary tests
- [ ] No security vulnerabilities
- [ ] Performance impact acceptable