- Implement content for 200+ TODO placeholders across all documentation - Create complete documentation structure: Getting-Started, User-Guide, Developer-Guide, Reference - Add comprehensive guides for commands, agents, modes, MCP servers, flags, session management - Implement technical architecture, contributing, testing, and security documentation - Create examples cookbook, troubleshooting guide, and best practices documentation - Update administrative files: CONTRIBUTING.md, SECURITY.md, PUBLISHING.md, CODE_OF_CONDUCT.md - Ensure factual accuracy based on actual SuperClaude implementation analysis - Maintain professional structure with progressive complexity and cross-references - Provide complete coverage from beginner to expert level usage 🤖 Generated with [Claude Code](https://claude.ai/code) Co-Authored-By: Claude <noreply@anthropic.com>
24 KiB
SuperClaude Session Management Guide 🧠
Introduction
SuperClaude transforms Claude Code from a stateless assistant into a persistent development partner through intelligent session management. Sessions preserve project context, accumulated insights, and development progress across interruptions, creating continuity that builds expertise over time.
Persistent Intelligence: Load project context with /sc:load, work naturally with accumulated understanding, save progress with /sc:save, and resume exactly where you left off. SuperClaude remembers your codebase structure, decisions made, patterns discovered, and project goals.
Cross-Session Learning: Each session builds on previous understanding, creating a development partner that becomes more effective with your specific project over time.
Session Fundamentals
What is a SuperClaude Session? A session is a persistent development context that includes:
- Project Understanding: Codebase structure, architecture patterns, dependencies
- Development History: Decisions made, problems solved, patterns applied
- Current State: Active tasks, progress tracking, next steps
- Learned Insights: Code conventions, team preferences, domain knowledge
Session Intelligence vs Standard Claude:
- Standard Claude: Each conversation starts fresh, no project memory
- SuperClaude Sessions: Cumulative understanding that builds over time
Session Types:
Project Sessions: Long-term development context for specific codebases
- Persistent across weeks/months of development
- Accumulates architectural understanding
- Remembers team conventions and decisions
Task Sessions: Focused context for specific features or problems
- Short to medium-term (hours to days)
- Maintains task-specific context and progress
- Integrates with project sessions for broader context
Learning Sessions: Educational context for understanding complex systems
- Preserves learning progress and insights
- Builds conceptual understanding over time
- Connects theoretical knowledge with practical application
Session Persistence Powered by Serena MCP:
- Semantic code understanding with symbol-level navigation
- Project memory that survives Claude Code restarts
- Intelligent context loading based on current work
- Cross-session insight accumulation and pattern recognition
Session Commands
/sc:load - Session Initialization
Purpose: Load project context and initialize persistent development session
Usage Patterns:
# Load existing project context
/sc:load src/
# Load with specific focus
/sc:load --focus architecture existing-project/
# Load previous session by name
/sc:load "payment-integration-session"
# Load with fresh analysis
/sc:load --refresh large-codebase/
What Happens During Load:
- Project Structure Analysis: Serena MCP scans codebase organization
- Context Restoration: Previous session memory and insights loaded
- Pattern Recognition: Code conventions and architecture patterns identified
- Dependency Mapping: Component relationships and data flow understood
- Session Initialization: Persistent context established for continued work
Load Modes:
Smart Load (Default):
/sc:load project-directory/
- Automatically detects project type and structure
- Loads relevant previous session context
- Optimizes analysis based on project size and complexity
Focused Load:
/sc:load --focus security payment-system/
/sc:load --focus performance api-endpoints/
/sc:load --focus architecture microservices/
- Specialized analysis for specific concerns
- Activates domain experts and relevant tools
- Loads context specific to focus area
Fresh Load:
/sc:load --refresh --comprehensive legacy-system/
- Re-analyzes project from scratch
- Updates understanding with latest changes
- Comprehensive re-indexing of codebase
Example Load Experience:
/sc:load e-commerce-platform/
→ 🔍 Analyzing project structure...
→ 📂 Detected: React frontend + Node.js API + PostgreSQL
→ 🧠 Loading previous session: "checkout-optimization-work"
→ 📝 Restored context: 847 files analyzed, 23 patterns identified
→ ✅ Session ready: Continue checkout flow optimization
→ 💡 Next: Complete payment validation implementation
/sc:save - Session Persistence
Purpose: Preserve session context and development progress for future continuation
Usage Patterns:
# Save current session with automatic naming
/sc:save
# Save with descriptive name
/sc:save "authentication-module-complete"
# Save checkpoint during long work
/sc:save --checkpoint "api-endpoints-analysis-done"
# Save with summary
/sc:save --description "Completed user registration flow, ready for testing"
What Gets Saved:
- Project Understanding: Current codebase analysis and insights
- Work Progress: Completed tasks, current focus, next steps
- Decision History: Architectural choices, patterns applied, trade-offs made
- Code Context: Modified files, dependencies, integration points
- Learning Insights: Discovered patterns, conventions, best practices
Save Types:
Milestone Save:
/sc:save "payment-integration-complete"
- Major feature completion or phase end
- Comprehensive context preservation
- Clear handoff point for future sessions
Checkpoint Save:
/sc:save --checkpoint "database-schema-analysis"
- Periodic progress preservation
- Intermediate work state capture
- Recovery point for complex work
Automatic Save:
# Triggered automatically during:
- Long-running tasks (every 30 minutes)
- Before risky operations
- Session interruption detection
- Context limit approaching
Save Output Example:
/sc:save "user-dashboard-feature-complete"
→ 💾 Saving session context...
→ 📊 Progress summary:
✅ 3 components implemented
✅ 12 tests passing
✅ API integration complete
📝 Documentation updated
→ 🧠 Context preserved:
- 247 files in working memory
- 15 architectural patterns identified
- 8 integration points mapped
→ ✅ Session saved: "user-dashboard-feature-complete"
→ 🔄 Ready for future continuation
/sc:reflect - Session Validation
Purpose: Analyze session progress, validate task completion, and assess development quality
Usage Patterns:
# General session reflection
/sc:reflect
# Specific task validation
/sc:reflect "payment-processing-implementation"
# Quality assessment
/sc:reflect --type quality-review
# Completion verification
/sc:reflect --type completion-check
Reflection Types:
Progress Analysis:
/sc:reflect --type progress
- Current task status and completion percentage
- Blockers and dependencies identified
- Next steps and priority recommendations
- Timeline and effort assessment
Quality Review:
/sc:reflect --type quality
- Code quality assessment and improvements needed
- Test coverage and validation gaps
- Security and performance considerations
- Documentation and maintainability review
Completion Verification:
/sc:reflect --type completion
- Task completion criteria assessment
- Integration and testing verification
- Deployment readiness evaluation
- Handoff preparation and next steps
Learning Assessment:
/sc:reflect --type learning
- Insights gained and patterns discovered
- Decision rationale and trade-offs made
- Knowledge gaps and learning opportunities
- Best practices and conventions established
Example Reflection Output:
/sc:reflect --type completion "user-authentication-system"
→ 🎯 Task Completion Analysis:
✅ JWT authentication implemented
✅ Password hashing and validation
✅ Session management
⚠️ Email verification pending
❌ OAuth integration incomplete
→ 📊 Quality Assessment:
✅ Security best practices followed
✅ 89% test coverage achieved
⚠️ Performance testing needed
✅ Documentation complete
→ 🔄 Next Steps:
1. Complete email verification flow
2. Implement OAuth providers
3. Performance optimization
4. Production deployment preparation
→ 💡 Session Status: 75% complete, 2-3 days remaining
Session Lifecycle Management
Complete Session Workflow:
Phase 1: Session Initialization
# 1. Project Discovery and Loading
/sc:load new-project/
→ Project analysis and context establishment
# 2. Goal Setting and Planning
/sc:brainstorm "improve user onboarding experience"
→ Requirements discovery and strategic planning
# 3. Initial Analysis
/sc:analyze --focus user-experience --think-hard
→ Comprehensive system understanding
Phase 2: Active Development
# 4. Implementation Planning
/sc:workflow "streamlined-onboarding-flow"
→ Systematic implementation strategy
# 5. Feature Development (Iterative)
/sc:implement "progressive-user-registration"
→ Coordinated development with automatic progress tracking
# 6. Regular Checkpoints
/sc:save --checkpoint "registration-flow-complete"
→ Periodic progress preservation
Phase 3: Validation and Quality
# 7. Progress Review
/sc:reflect --type progress
→ Status assessment and next steps
# 8. Quality Assurance
/sc:test --comprehensive user-onboarding/
→ Testing and validation workflows
# 9. Quality Review
/sc:reflect --type quality
→ Code review and improvement identification
Phase 4: Completion and Handoff
# 10. Completion Verification
/sc:reflect --type completion
→ Task completion criteria validation
# 11. Final Documentation
/sc:document --type user-guide onboarding-system/
→ Documentation and knowledge transfer
# 12. Session Archive
/sc:save "onboarding-improvement-complete"
→ Final context preservation and handoff preparation
Session State Transitions:
Active Session States:
- Discovery: Requirements exploration and goal setting
- Analysis: System understanding and strategy development
- Implementation: Active development and feature creation
- Validation: Testing, review, and quality assurance
- Completion: Final verification and handoff preparation
Session Continuity Patterns:
- Daily Sessions: Load → Work → Checkpoint → Save
- Weekly Sprints: Load → Sprint planning → Daily work → Sprint review → Save
- Feature Cycles: Load → Analysis → Implementation → Testing → Completion → Save
Memory and Context
Serena MCP-Powered Memory System:
Project Memory Components:
- Structural Memory: Codebase organization, file relationships, dependencies
- Semantic Memory: Code meaning, business logic, domain concepts
- Historical Memory: Development decisions, evolution patterns, change history
- Contextual Memory: Working sessions, current focus, progress state
Memory Operations:
Memory Accumulation:
# As you work, memory automatically builds:
/sc:analyze user-service/
→ Records: Service patterns, data flows, API contracts
/sc:implement "user-preferences"
→ Records: Implementation patterns, coding conventions, integration points
/sc:troubleshoot "performance-issue"
→ Records: Problem patterns, solution strategies, optimization techniques
Memory Recall:
# Previous insights automatically inform current work:
/sc:implement "user-notifications"
→ Recalls: User service patterns, preference storage, communication flows
→ Applies: Established conventions, tested patterns, integration strategies
Memory Types:
Architectural Memory:
{
"patterns": ["MVC", "Repository Pattern", "Dependency Injection"],
"conventions": ["camelCase variables", "async/await preferred"],
"decisions": ["PostgreSQL for persistence", "JWT for authentication"],
"integrations": ["Stripe API", "SendGrid", "Redis cache"]
}
Development Memory:
{
"workflows": ["TDD approach", "Feature branch strategy"],
"quality_gates": ["ESLint rules", "Test coverage >80%"],
"preferences": ["Functional components", "Hooks over classes"],
"tooling": ["Jest testing", "Prettier formatting"]
}
Context Preservation:
{
"current_task": "implement-user-dashboard",
"progress": "components-complete-api-pending",
"next_steps": ["API integration", "state management", "testing"],
"blockers": ["API rate limiting", "design system tokens"]
}
Memory Intelligence:
Smart Context Loading:
- Loads relevant memory based on current work
- Prioritizes recent and frequently accessed patterns
- Connects related concepts and components
Pattern Recognition:
- Identifies recurring code patterns and conventions
- Suggests consistent approaches based on project history
- Detects deviations from established patterns
Predictive Context:
- Anticipates needed information based on current task
- Pre-loads related components and dependencies
- Suggests next steps based on similar previous work
Session Workflows
Common Session Patterns:
Daily Development Session
# Morning: Resume previous work
/sc:load project/
→ Context restoration and progress review
# Planning: Review and plan day's work
/sc:reflect --type progress
→ Status assessment and priority setting
# Work: Active development with periodic saves
/sc:implement "current-feature"
/sc:save --checkpoint "feature-milestone"
# Evening: End-of-day preservation
/sc:reflect --type completion
/sc:save "end-of-day-progress"
Feature Development Cycle
# Discovery Phase
/sc:load project/
/sc:brainstorm "new-feature-requirements"
/sc:save --checkpoint "requirements-complete"
# Planning Phase
/sc:workflow "feature-implementation-plan"
/sc:design "feature-architecture"
/sc:save --checkpoint "planning-complete"
# Implementation Phase
/sc:implement "feature-core-functionality"
/sc:save --checkpoint "core-complete"
/sc:implement "feature-integration"
/sc:save --checkpoint "integration-complete"
# Validation Phase
/sc:test --comprehensive feature/
/sc:reflect --type quality
/sc:save "feature-complete"
Bug Investigation Session
# Load with focus on problem area
/sc:load --focus debugging problematic-module/
# Systematic investigation
/sc:troubleshoot "specific-bug-description"
/sc:analyze --focus root-cause affected-components/
# Solution implementation
/sc:implement "bug-fix-solution"
/sc:test --focus regression bug-area/
# Validation and documentation
/sc:reflect --type completion
/sc:document --type bug-report "issue-resolution"
/sc:save "bug-fix-complete"
Learning and Exploration Session
# Focused learning context
/sc:load --focus architecture new-codebase/
# Systematic exploration
/sc:analyze --introspect --think-hard codebase-structure/
/sc:explain "complex-architectural-patterns"
# Knowledge consolidation
/sc:reflect --type learning
/sc:document --type architecture-notes findings/
/sc:save "architecture-understanding-session"
Code Review Session
# Load changes context
/sc:load --focus quality pending-changes/
# Comprehensive review
/sc:analyze --focus quality --think-hard changes/
/sc:test --focus regression change-areas/
# Quality assessment
/sc:reflect --type quality
/sc:document --type review-notes quality-assessment/
/sc:save "code-review-complete"
Session Optimization Patterns:
Short Sessions (1-2 hours):
- Quick load with specific focus
- Targeted work on single component
- Checkpoint saves for continuity
Medium Sessions (Half-day):
- Comprehensive load and planning
- Multi-component development
- Regular checkpoints and quality reviews
Long Sessions (Full-day):
- Full context loading and analysis
- Complex feature development cycles
- Multiple validation and reflection points
Multi-Session Projects
Long-Term Project Management:
Project Session Architecture
# Master Project Session
/sc:load enterprise-platform/
→ Maintains overall project context and architecture understanding
# Feature Branch Sessions
/sc:load --branch feature/user-management user-service/
/sc:load --branch feature/payment-integration payment-service/
→ Focused context for specific feature development
# Integration Sessions
/sc:load --integration-focus platform-services/
→ Cross-service integration and system-level concerns
Session Hierarchy Management
Project Level (Months):
- Overall architecture and system understanding
- Cross-cutting concerns and integration patterns
- Long-term technical decisions and evolution
Epic Level (Weeks):
- Feature set implementation and integration
- Domain-specific patterns and conventions
- Epic-level progress and quality tracking
Story Level (Days):
- Individual feature implementation
- Component-level development and testing
- Story completion and handoff
Session Coordination Patterns:
Team Coordination:
# Shared Project Context
/sc:load --shared team-project/
→ Common understanding accessible to all team members
# Individual Developer Sessions
/sc:load --developer alice team-project/user-auth/
/sc:load --developer bob team-project/payment-system/
→ Personal development context within shared project
# Integration Sessions
/sc:load --integration team-project/
→ Cross-developer integration and system-level work
Cross-Session Continuity:
Session Handoff:
# End of developer session
/sc:save --handoff "alice-user-auth-complete" --next-developer bob
# New developer pickup
/sc:load --handoff "alice-user-auth-complete"
→ Complete context transfer with work continuation
Progress Synchronization:
# Daily standup preparation
/sc:reflect --type team-progress
→ Team-level progress summary and coordination
# Sprint planning context
/sc:load --sprint-context team-project/
→ Sprint-level understanding and planning context
Long-Term Memory Evolution:
Memory Consolidation:
- Weekly: Consolidate daily insights into persistent patterns
- Monthly: Archive completed features, preserve key learnings
- Quarterly: Architectural review and pattern evolution
Memory Inheritance:
- New features inherit patterns from completed work
- Team members inherit shared conventions and decisions
- Project evolution builds on accumulated architectural understanding
Session Troubleshooting
Common Session Issues:
Session Context Lost
Problem: /sc:load doesn't restore expected context
# Symptoms
/sc:load project/
→ "No previous session found" or minimal context restored
Solutions:
# 1. Check available sessions
/sc:load --list-sessions
# 2. Load by specific session name
/sc:load "session-name"
# 3. Fresh analysis if needed
/sc:load --refresh --comprehensive project/
# 4. Check Serena MCP status
SuperClaude status --mcp serena
Memory Fragmentation
Problem: Session memory becomes inconsistent or incomplete
# Symptoms
- Conflicting pattern suggestions
- Missing context for recent work
- Inconsistent architectural understanding
Solutions:
# 1. Memory consolidation
/sc:reflect --type memory-health
/sc:save --consolidate "clean-session-state"
# 2. Selective memory refresh
/sc:load --refresh-memory specific-area/
# 3. Clean session restart
/sc:save --archive "old-session"
/sc:load --fresh project/
Session Performance Issues
Problem: Slow session loading or excessive memory usage
# Symptoms
- Long load times (>30 seconds)
- Memory warnings during operation
- Reduced response quality
Solutions:
# 1. Optimize session scope
/sc:load --scope module target-area/
# 2. Use focused loading
/sc:load --focus performance specific-concern/
# 3. Memory cleanup
/sc:save --cleanup "optimized-session"
# 4. Check system resources
SuperClaude diagnose --memory
Integration Conflicts
Problem: Multi-session project coordination issues
# Symptoms
- Conflicting architectural decisions
- Integration pattern mismatches
- Cross-team context confusion
Solutions:
# 1. Sync with master session
/sc:load --sync-master project/
# 2. Resolve conflicts explicitly
/sc:reflect --type integration-conflicts
/sc:resolve-conflicts --strategy team-lead-wins
# 3. Re-establish shared context
/sc:load --shared-refresh team-project/
Session Recovery Procedures:
Emergency Session Recovery
# 1. Check session backups
/sc:load --list-backups
# 2. Restore from backup
/sc:load --restore-backup "backup-timestamp"
# 3. Partial recovery if needed
/sc:load --partial-restore specific-components/
# 4. Rebuild from artifacts
/sc:load --rebuild-from git-history project/
Session Health Monitoring
# Regular health checks
/sc:reflect --type session-health
→ Memory consistency, context completeness, performance metrics
# Memory optimization
/sc:save --optimize "clean-session"
→ Consolidate insights, remove redundant context
# Performance tuning
/sc:load --performance-mode project/
→ Optimized loading for resource-constrained environments
Best Practices for Session Reliability:
- Regular Saves: Save at natural breakpoints and end of work
- Descriptive Names: Use clear, descriptive session names
- Health Monitoring: Regular reflection and health checks
- Backup Strategy: Multiple checkpoint saves during long work
- Scope Management: Load only necessary context for current work
- Memory Hygiene: Periodic consolidation and cleanup
Related Guides
Learning Progression:
🌱 Essential (Week 1)
- Quick Start Guide - First session experience
- Commands Reference - Session commands (/sc:load, /sc:save, /sc:reflect)
- Installation Guide - Serena MCP setup for sessions
🌿 Intermediate (Week 2-3)
- MCP Servers Guide - Serena MCP for persistent memory
- Behavioral Modes - Task Management mode for session coordination
- Agents Guide - Agent persistence across sessions
🌲 Advanced (Month 2+)
- Best Practices - Session optimization strategies
- Examples Cookbook - Multi-session project patterns
- Flags Guide - Session control flags and optimization
🔧 Expert
- Technical Architecture - Session implementation details
- Contributing Code - Extending session capabilities
Session-Specific Learning:
🎯 Session Command Mastery:
- Load Patterns:
/sc:loadvariations for different contexts - Save Strategies: Checkpoint vs milestone vs completion saves
- Reflection Types: Progress, quality, completion, learning assessments
🧠 Memory Management:
- Context Building: How memory accumulates and evolves
- Pattern Recognition: Understanding session intelligence
- Memory Optimization: Efficient context management
👥 Team Coordination:
- Shared Sessions: Multi-developer session coordination
- Handoff Patterns: Session transfer and continuity
- Integration Sessions: Cross-team collaboration
🔄 Project Lifecycle:
- Daily Workflows: Effective day-to-day session patterns
- Feature Cycles: Multi-session feature development
- Long-term Projects: Enterprise-scale session management
💡 Session Mastery Tips:
- Start Small: Begin with simple load-work-save patterns
- Build Habits: Regular reflection and checkpoint saves
- Learn Memory: Understand how context builds over time
- Optimize Gradually: Advanced patterns as projects grow
- Monitor Health: Regular session health and performance checks
Common Session Workflows:
- New Project: Load → Analyze → Plan → Save
- Daily Work: Load → Progress check → Work → Checkpoint → Save
- Feature Development: Load → Brainstorm → Implement → Test → Reflect → Save
- Bug Investigation: Load focused → Troubleshoot → Fix → Validate → Save
- Code Review: Load changes → Analyze quality → Review → Document → Save