Store Migration Workflow

Store Migration Workflow

What This Skill Does

Orchestrates the migration from monolithic state stores to domain-driven stores using:

• GOAP Planning: A* pathfinding for optimal action sequences
• Hive-Mind Coordination: Collective intelligence for complex refactoring
• Neural Pattern Learning: Train on successful migration patterns
• AgentDB Memory: Persist migration state across sessions

Prerequisites

• Claude Flow v2.0+ (npx claude-flow@alpha)
• Zustand-based state management
• Domain store targets identified

Quick Start

# 1. Initialize hive-mind
npx claude-flow hive-mind init

# 2. Store GOAP world state
# Current: {monolithExists: true, domainStoresCreated: false}
# Goal: {monolithExists: false, domainStoresCreated: true, testsPass: true}

# 3. Generate action plan
# Actions: create_domain_stores → migrate_imports → deprecate_monolith → run_tests → commit

GOAP Action Definitions

Available Actions

Example GOAP Plan

{
  "currentState": {
    "monolithExists": true,
    "monolithExported": true,
    "domainStoresCreated": true,
    "importsUpdated": true,
    "testsPass": "unknown"
  },
  "goalState": {
    "monolithExists": false,
    "testsPass": true,
    "migrationComplete": true
  },
  "optimalPath": [
    "remove_export",
    "deprecate_monolith",
    "run_tests",
    "commit_changes"
  ],
  "totalCost": 7
}

Hive-Mind Integration

# Spawn refactoring swarm
npx claude-flow hive-mind spawn "Migrate gameStore to domain stores" \
  --queen-type tactical \
  --max-workers 5 \
  --consensus weighted

Worker Specialization

• Analyzer Agent: Identify store responsibilities
• Coder Agent: Create domain stores
• Migrator Agent: Update imports
• Tester Agent: Verify no regressions
• Reviewer Agent: Validate architecture

Neural Pattern Training

# Train on migration patterns
npx claude-flow neural train \
  --pattern-type optimization \
  --data '{"domain":"store-migration","patterns":["decompose","migrate","verify"]}'

Learned Patterns

1. decompose-by-domain: Split by business domain, not technical layer
2. migrate-incrementally: One import at a time with tests
3. deprecate-not-delete: Add JSDoc deprecation before removal
4. verify-continuously: Run tests after each migration step

Memory Persistence

// Store migration state
await memory.store('goap/world-state', currentState, 'goap');
await memory.store('goap/goal-state', goalState, 'goap');
await memory.store('goap/action-log/*', actionResults, 'goap');

// Retrieve for session resume
const worldState = await memory.retrieve('goap/world-state');

Step-by-Step Migration

Phase 1: Analysis

1. Identify god object (>500 LOC, >50 methods)
2. Map responsibilities to domains
3. Create domain store interfaces

Phase 2: Creation

1. Create domain stores with single responsibility
2. Add store coordinator for cross-store communication
3. Export from barrel file

Phase 3: Migration

1. Find all monolith imports (grep -r "useGameStore")
2. Replace with domain store imports
3. Update component selectors

Phase 4: Cleanup

1. Remove monolith export from index.ts
2. Add deprecation JSDoc to monolith
3. Run full test suite
4. Commit with migration message

Phase 5: Deletion (Future)

1. Verify zero imports remain
2. Delete monolith file
3. Remove from git history (optional)

Replanning Triggers

Success Criteria

• All imports migrated to domain stores
• Monolith export removed
• Deprecation JSDoc added
• All tests pass
• TypeScript compiles
• Lint passes
• Architecture grade improved

Troubleshooting

Issue: Circular Dependencies

Solution: Use store coordinator pattern with Zustand subscriptions

Issue: Type Mismatches

Solution: Create shared type definitions in @/types

Issue: Test Failures After Migration

Solution: Update test mocks to use domain stores

Related Skills

• agentdb-memory-patterns - Persistent memory
• hive-mind-advanced - Collective coordination
• goap-planning - Action planning
• sparc-methodology - Structured development

References

• Architecture Assessment
• Zustand Best Practices
• GOAP Algorithm

Created: 2025-12-03 Version: 1.0.0 Category: Architecture Refactoring Difficulty: Intermediate