Workflow Patterns
Common patterns and strategies for effectively leading your AI development team
Workflow Patterns
Learn proven patterns for orchestrating your AI team effectively. These workflows help you maximize productivity and quality.
Feature Development Pattern
The most common workflow for adding new features:
1. Architecture First
> Design a user notification system that supports email, SMS, and in-app notifications
The Architect Agent will:
- Design the system architecture
- Recommend design patterns
- Identify required components
- Suggest technology choices
2. Incremental Implementation
> Implement the notification service base class with the architecture you designed
3. Parallel Development
> Create email, SMS, and in-app notification providers based on the base class
Multiple Developer Agents work simultaneously on different providers.
4. Testing & Documentation
> Write comprehensive tests for all notification providers and create API documentation
Bug Fixing Pattern
Efficient approach to identifying and fixing issues:
1. Describe the Problem
> Users report that emails are not being sent when they reset their password
2. Investigation
The team will:
- Analyze the code flow
- Identify potential issues
- Suggest debugging steps
3. Fix Implementation
> Fix the email sending issue in the password reset flow
4. Verification
> Add tests to ensure password reset emails are sent correctly
Refactoring Pattern
Improving existing code without changing functionality:
1. Analysis Request
> Analyze the user service for potential improvements and technical debt
2. Refactoring Plan
> Create a refactoring plan to improve the user service code quality
3. Incremental Refactoring
> Refactor the user authentication methods to use async/await
4. Testing
> Ensure all tests still pass after refactoring
Greenfield Project Pattern
Starting a new project from scratch:
1. Project Setup
> Create a new Express.js API project with TypeScript, ESLint, and Jest
2. Core Architecture
> Design the architecture for a task management API with users, projects, and tasks
3. Database Design
> Create the database schema for the task management system
4. Implementation Phases
> Implement user authentication and registration
> Add project CRUD operations with authorization
> Create task management endpoints
> Add real-time updates using WebSockets
Code Review Pattern
Getting AI assistance with code review:
1. Request Review
> Review this function for potential improvements: [paste code]
2. Security Audit
> Check this authentication code for security vulnerabilities
3. Performance Analysis
> Analyze this database query for performance issues
Documentation Pattern
Creating comprehensive documentation:
1. API Documentation
> Generate OpenAPI/Swagger documentation for all endpoints
2. Setup Guides
> Create a setup guide for new developers joining the project
3. Architecture Documentation
> Document the system architecture with diagrams and explanations
Learning Pattern
Using xSwarm to learn new technologies:
1. Exploration
> Explain how GraphQL differs from REST and when to use each
2. Implementation
> Show me how to implement a simple GraphQL server with type safety
3. Best Practices
> What are the best practices for GraphQL schema design?
Integration Pattern
Adding third-party services:
1. Research
> Compare Stripe, PayPal, and Square for payment processing
2. Implementation Plan
> Design the payment integration architecture using Stripe
3. Step-by-Step Integration
> Implement Stripe checkout for subscription payments
> Add webhook handling for payment events
> Create payment history and invoice generation
Debugging Pattern
Systematic approach to fixing issues:
1. Error Analysis
> This error appears in production: [error message]. Help me understand what's happening
2. Reproduction
> Create a test case that reproduces this error
3. Solution
> Fix the error and add proper error handling
Migration Pattern
Upgrading or migrating systems:
1. Analysis
> Analyze the effort required to migrate from Express to Fastify
2. Migration Plan
> Create a step-by-step migration plan with minimal downtime
3. Incremental Migration
> Migrate the authentication module to Fastify
Best Practices for All Patterns
- Start Small: Break large tasks into smaller, manageable pieces
- Iterate Often: Get feedback early and adjust
- Maintain Context: Reference previous work for consistency
- Review Output: Always review AI-generated code
- Test Everything: Ensure comprehensive test coverage
Next Steps
- Explore Advanced Features
- See Real-World Examples
- Learn Troubleshooting Tips