-
-
Notifications
You must be signed in to change notification settings - Fork 34
Basic Workflow
Understanding the RIPER workflow is key to maximizing your productivity with AI assistance. This guide explains how to use the framework effectively.
RIPER stands for:
- Research - Understand the problem
- Innovate - Explore solutions
- Plan - Design the approach
- Execute - Implement the solution
- Review - Validate the results
Each mode has specific purposes and restrictions to keep you and the AI focused.
┌─────────────┐
│ 🔍 RESEARCH │ ← Start here
└──────┬──────┘
↓
┌─────────────┐
│ 💡 INNOVATE │
└──────┬──────┘
↓
┌─────────────┐
│ 📝 PLAN │
└──────┬──────┘
↓
┌─────────────┐
│ ⚙️ EXECUTE │
└──────┬──────┘
↓
┌─────────────┐
│ 🔎 REVIEW │
└─────────────┘
Purpose: Gather information and understand requirements
- ✅ Read existing code and documentation
- ✅ Ask questions about the project
- ✅ Analyze current implementation
- ✅ Document findings
- ❌ Write or modify code
- ❌ Create new files
- ❌ Make design decisions
/research
"Analyze the current authentication system and identify security vulnerabilities"
!af src/auth/login.js # Add relevant files to context
!ad src/auth/ # Add authentication folder
- 📋
projectbrief.md- Requirements documented - 💻
techContext.md- Technical findings recorded
Purpose: Explore ideas and potential solutions
- ✅ Suggest new approaches
- ✅ Explore alternatives
- ✅ Discuss pros and cons
- ✅ Create conceptual designs
- ❌ Implement solutions
- ❌ Write actual code
- ❌ Make final decisions
/innovate
"What are modern alternatives to session-based authentication?"
!ac authenticationConcepts # Track conceptual ideas
!adoc OAuth2 Specification # Reference relevant docs
- 🔮
activeContext.md- Ideas and concepts tracked - 🏛️
systemPatterns.md- Potential patterns noted
Purpose: Create detailed implementation plans
- ✅ Create specifications
- ✅ Design architecture
- ✅ Sequence implementation steps
- ✅ Define success criteria
- ❌ Write production code
- ❌ Execute the plan
- ❌ Skip planning steps
/plan
"Create a detailed plan to migrate from sessions to JWT authentication"
# The AI will create a numbered checklist:
1. Install required dependencies (jsonwebtoken, bcrypt)
2. Create JWT utility functions
3. Modify user model for refresh tokens
4. Update login endpoint
5. Create token refresh endpoint
6. Update middleware for JWT validation
7. Migrate existing sessions
8. Update client-side token handling
!cp Identify critical sections that need protection
!cg Mark sections that require review before changes
- 🔮
activeContext.md- Plan details stored - 📊
progress.md- Milestones defined
Purpose: Implement exactly according to plan
- ✅ Write code following the plan
- ✅ Create and modify files
- ✅ Run tests
- ✅ Fix bugs
- ❌ Deviate from the plan
- ❌ Add unplanned features
- ❌ Search web (stay focused!)
- ❌ Redesign on the fly
/execute
"Implement step 2 from our plan: Create JWT utility functions"
# AI writes code exactly as planned
!af src/utils/jwt.js # Track new files
!ac generateToken() # Track new functions
// !cp PROTECTED - JWT Secret
const JWT_SECRET = process.env.JWT_SECRET;
// !cp END-P
// !cc CRITICAL - Token Generation
function generateToken(user) {
return jwt.sign(
{ id: user.id, email: user.email },
JWT_SECRET,
{ expiresIn: '24h' }
);
}
// !cc END-C- 📊
progress.md- Implementation progress tracked - 🛡️
protection.md- Protected regions recorded
Purpose: Validate implementation against requirements
- ✅ Check code quality
- ✅ Verify against plan
- ✅ Identify issues
- ✅ Suggest improvements
- ❌ Modify code
- ❌ Implement fixes
- ❌ Change requirements
/review
"Review the JWT implementation against our security requirements"
!ac generateToken() # Review specific functions
!af test/auth.test.js # Check test coverage
- ✅ PASS: Matches plan exactly
⚠️ WARNING: Minor deviations found- ❌ FAIL: Does not match plan
- 📊
progress.md- Review results recorded - 🔮
activeContext.md- Issues documented
/r → /i → /p → /e → /rev
-
Research → Innovate
- When you understand the problem
- Requirements are clear
- Ready to explore solutions
-
Innovate → Plan
- When you've chosen an approach
- Ready to create specifications
- Need concrete steps
-
Plan → Execute
- When plan is complete
- All steps are clear
- Ready to implement
-
Execute → Review
- When implementation is done
- Need to validate work
- Ready for quality check
You can jump to any mode when needed:
/review → /research # Need more information
/execute → /plan # Plan needs revision
/innovate → /execute # Emergency fix (not recommended)
!ad src/ # Broad context
!adoc README.md # Documentation
!ag main # Git history
!ac conceptA() # Conceptual references
!an "Design Ideas" # Notepad references
!adoc "Best Practices" # External docs
!af config.json # Specific files
!ar coding-standards # Project rules
!ad src/components/ # Target directories
!pf critical.js # Pin critical files
!ac implementedFunc() # Track new code
!cc # Clear old context
!ag feature-branch # Git differences
!af test-results.log # Test outputs
!ac reviewTarget() # Specific functions
-
During Planning:
- Identify sensitive areas
- Mark future protection needs
-
During Execution:
- Protect as you write
- Guard critical logic
-
During Review:
- Verify protection coverage
- Add missing protections
// Research: Identify sensitive areas
// Plan: Design protection strategy
// Execute: Implement with protection
// !cp PROTECTED - Never modify
const CRITICAL_CONFIG = {
apiKey: process.env.API_KEY,
secretKey: process.env.SECRET_KEY
};
// !cp END-P
// Review: Verify all critical code is protected!ckp # Check current mode permissions
!pm write_code # Can I write code now?
!vm read_files # What mode for reading files?
-
"Cannot write code in RESEARCH mode"
- Switch to EXECUTE mode first
-
"Cannot search web in EXECUTE mode"
- Complete execution, then switch to RESEARCH
-
"Cannot modify plan in EXECUTE mode"
- Switch back to PLAN mode
Don't skip steps. Each mode builds on the previous.
Your progress is saved automatically. Reference it:
"According to our plan in progress.md..."
"As documented in projectbrief.md..."
- Research: Broad context
- Plan: Specific targets
- Execute: Narrow focus
- Review: Comparison context
Add protection during planning and execution, not as an afterthought.
The restrictions exist to improve outcomes. Embrace them!
-
/r- Research existing code -
/i- Explore feature approaches -
/p- Plan implementation -
/e- Build feature -
/rev- Validate completion
-
/r- Investigate bug -
/p- Plan fix approach -
/e- Implement fix -
/rev- Verify fix works
-
/r- Analyze current code -
/i- Explore refactoring options -
/p- Plan refactoring steps -
/e- Execute refactoring -
/rev- Ensure functionality preserved
Work on multiple features by switching context:
!cc # Clear context
!cm # Set mode-appropriate context
!af feature-b.js # Focus on different feature
/r → /i → /p → /e → /rev # Quick mode switches
!cm # Auto-set context for mode
For urgent fixes (use sparingly):
/p # Quick plan
"Emergency fix for [issue]"
/e # Immediate execution
Now that you understand the basic workflow:
- 🏗️ Memory System Deep Dive
- 🛡️ Advanced Protection Strategies
- 📎 Context Management Mastery
- 🔧 Troubleshooting Workflows
- 🔄 Mode Transitions
- 💾 Memory Management
- 🛡️ Protection Workflow
- 📎 Context Management
- 👥 Team Collaboration
-
Installation Issues
- Node.js Version Compatibility
- Package Installation Failures
- Framework Dependencies Missing
- Database Connection Issues
- Port Conflicts
- Environment Setup Issues
- Build and Development Issues
- Framework CLI Issues
-
Configuration & Runtime Issues
- Framework Configuration Problems
- Runtime Performance Issues
- Module Loading and Plugin Issues
- Database and Storage Issues
- Memory Leaks and High Memory Usage
- High CPU Usage
-
BMAD Module Issues
- BMAD Module Initialization Problems
- Business Model Canvas Issues
- Stakeholder Management Issues
- Analytics and Reporting Issues
- Performance Optimization
-
Database & API Issues
- Database Connection Problems
- Database Migration Issues
- API Performance and Reliability Issues
- Data Consistency Issues
- Transaction Problems
-
Performance & Memory Issues
- Memory Management
- CPU Optimization
- Database Query Performance
- Caching Issues
- Resource Monitoring
-
Security & Authentication Issues
- Authentication Failures
- Authorization Problems
- JWT Token Issues
- Session Management
- CORS and Security Headers
- SSL/TLS Configuration
-
Deployment & Production Issues
- Production Deployment Failures
- Environment Configuration
- Load Balancing Issues
- Monitoring and Logging
- Backup and Recovery
When reporting issues, please include:
- Framework version (
npm list @cursoriper/core) - Node.js version (
node --version) - Operating system and version
- Error messages and stack traces
- Steps to reproduce the issue
- Configuration files (sanitized)
- Recent changes or deployments
- Technical Support: support@cursoriper.com
- Documentation: https://docs.cursoriper.com
- Community Forum: https://community.cursoriper.com