# Feature Addition Workflow workflow: id: feature-addition name: Add Feature to Existing Project description: Works for both greenfield and brownfield projects duration: 2-10 days (based on size) token_budget: 4500 (small) | 8000 (medium) | 12000 (large) # SIZE ESTIMATION (30 minutes) sizing: agent: analyst + architect duration: 30 minutes tokens: 500 size_definitions: small: effort: 1-2 days files: 1-3 files changed complexity: Using existing patterns examples: - Add new field to form - Create new page with CRUD - Add filter to list - Update UI component medium: effort: 2-5 days files: 4-10 files changed complexity: Some new patterns examples: - Payment integration - Email notifications - File upload system - Advanced filtering large: effort: 5-10 days files: 10+ files changed complexity: New architecture examples: - Multi-step wizard - Real-time collaboration - Advanced analytics - Complex workflow decision: small: Skip to implementation medium: Do quick design large: Full design phase # SMALL FEATURE (Skip most phases) small_feature: duration: 1-2 days tokens: 4500 quick_analysis: agent: appropriate-developer duration: 1 hour outputs: - feature-brief.md (1 page max) checklist: - [ ] What changes are needed? - [ ] Which files to modify? - [ ] Any API changes? - [ ] Test strategy? implementation: agent: appropriate-developer duration: 4-12 hours tasks: - Make code changes - Add/update tests - Update documentation - Manual testing outputs: - Working feature - Updated tests deployment: duration: 1 hour tasks: - Code review - Merge to main - Deploy to production # MEDIUM FEATURE medium_feature: duration: 2-5 days tokens: 8000 # Phase 1: Analysis (2-4 hours) analysis: agent: analyst + architect duration: 2-4 hours tokens: 1000 outputs: - feature-requirements.md - impact-analysis.md analyze: - Detailed requirements - Affected components - API changes needed - Database schema updates - Integration points - Risks and dependencies # Phase 2: Design (2-4 hours) design: agent: appropriate-specialist duration: 2-4 hours tokens: 1500 outputs: - technical-design.md - api-changes.yaml (if backend) - component-plan.md (if frontend) design_elements: - API contracts - Component structure - State management - Database changes - Error handling checkpoint: name: feature-design-checkpoint.md max_tokens: 300 # Phase 3: Implementation (2-5 days) implementation: agent: developers duration: 2-5 days tokens: 4000 parallel_tracks: frontend: - Build UI components - Implement state logic - Add form validation - Style components backend: - Create API endpoints - Update database schema - Add business logic - Handle errors testing: - Unit tests - Integration tests - Manual QA checkpoints: frequency: Daily max_tokens: 200 each # Phase 4: Testing & Deployment (4-8 hours) release: duration: 4-8 hours tokens: 1500 tasks: - Code review - Final QA - Deploy to staging - Smoke test - Deploy to production - Monitor metrics # LARGE FEATURE large_feature: duration: 5-10 days tokens: 12000 # Phase 1: Requirements (4-8 hours) requirements: agent: analyst duration: 4-8 hours tokens: 1500 outputs: - detailed-requirements.md - user-stories.md (5-10 stories) - acceptance-criteria.md - success-metrics.md # Phase 2: Architecture Review (4 hours) architecture: agent: js-solution-architect duration: 4 hours tokens: 2000 load: jit: - IF new_patterns: architecture-patterns.md - IF security: security-guidelines.md - IF complex_api: api-implementation-patterns.md outputs: - architecture-impact.md - integration-plan.md - risk-mitigation.md checkpoint: name: architecture-impact-checkpoint.md max_tokens: 400 # Phase 3: Design (1 day) design: agent: team-leads duration: 1 day tokens: 2500 parallel_design: frontend: - Component hierarchy - State architecture - Routing changes - UI/UX flow backend: - API specification - Database schema - Service architecture - Integration design infrastructure: - Performance requirements - Scaling strategy - Monitoring plan outputs: - frontend-design.md - backend-design.md - infrastructure-plan.md checkpoint: name: detailed-design-checkpoint.md max_tokens: 500 # Phase 4: Story Creation (2 hours) stories: agent: scrum-master duration: 2 hours tokens: 1000 outputs: - implementation-stories.md (5-10 stories) - sprint-plan.md story_types: - Database migrations - Backend API endpoints - Frontend components - Integration work - Testing & QA # Phase 5: Implementation (5-10 days) implementation: duration: 5-10 days tokens: 4000 parallel_teams: ui_team: - Build component library - Implement pages - Add interactions api_team: - Create endpoints - Add validation - Implement business logic data_team: - Schema migrations - Data transformations - Database optimization qa_team: - Write test cases - Automated testing - Manual QA daily_checkpoints: - Progress report - Blocker resolution - Integration status # Phase 6: Integration & Release (1-2 days) release: duration: 1-2 days tokens: 1000 phases: integration: - Merge feature branches - Integration testing - Performance testing - Security review deployment: - Deploy to staging - User acceptance testing - Production deployment - Post-deployment monitoring # FEATURE-SPECIFIC PATTERNS patterns: authentication_feature: components: - Login/register forms - Auth context/provider - Protected routes - Session management backend: - Auth endpoints - JWT middleware - Password hashing - Refresh token logic payment_integration: components: - Checkout flow - Payment forms - Order confirmation backend: - Stripe/payment integration - Webhook handlers - Order processing - Receipt generation real_time_feature: components: - WebSocket client - Real-time updates - Optimistic UI backend: - WebSocket server - Event broadcasting - State synchronization file_upload: components: - Upload UI - Progress tracking - File preview backend: - Upload endpoints - S3/storage integration - File validation # COMMON DECISIONS decisions: new_dependencies: evaluate: - Bundle size impact - Maintenance status - License compatibility - Security record approval: Tech lead required database_changes: approach: - Write migration scripts - Test rollback - Plan zero-downtime deployment backup: Always before schema changes breaking_changes: process: - Document clearly - Version API - Provide migration guide - Deprecation period # SUCCESS CRITERIA success_metrics: small: - Feature works as expected - Tests pass - Code reviewed - Deployed to production medium: - All acceptance criteria met - Test coverage adequate - Performance acceptable - Documentation updated large: - All stories complete - Full test coverage - Performance validated - User acceptance passed - Monitoring in place # DELIVERABLES deliverables: code: - [ ] Feature implementation - [ ] Unit tests - [ ] Integration tests - [ ] Updated types/interfaces documentation: - [ ] Feature documentation - [ ] API documentation (if applicable) - [ ] Migration guide (if breaking) - [ ] Updated README deployment: - [ ] Deployed to production - [ ] Monitoring configured - [ ] Rollback plan documented # QUICK REFERENCE quick_start: step_1: Estimate feature size (small/medium/large) step_2: Follow appropriate workflow path step_3: Use parallel development when possible step_4: Checkpoint daily for medium/large features step_5: Deploy incrementally if possible