Claude-Code-Workflow/.claude/commands/task/breakdown.md

name: task-breakdown description: Intelligent task decomposition with context-aware subtask generation usage: /task:breakdown [--strategy=<auto|interactive>] [--depth=<1-3>] argument-hint: task-id [optional: strategy and depth] examples: - /task:breakdown IMPL-1 - /task:breakdown IMPL-1 --strategy=auto - /task:breakdown IMPL-1.1 --depth=2 --strategy=interactive

Task Breakdown Command (/task:breakdown)

Overview

Intelligently breaks down complex tasks into manageable subtasks with automatic context distribution and agent assignment.

Core Principles

System: @/.claude/workflows/unified-workflow-system-principles.md
Task Schema: @/.claude/workflows/task-management-principles.md

Features

⚠️ CRITICAL: Before breakdown, MUST check for existing active session to avoid creating duplicate sessions.

Session Check Process

1. Check Active Session: Check for .workflow/.active-* marker file to identify active session containing the parent task.
2. Session Validation: Use existing active session containing the parent task
3. Context Integration: Load existing session state and task hierarchy

Smart Decomposition

• Auto Strategy: AI-powered subtask generation based on title
• Interactive Mode: Guided breakdown with suggestions
• Context Distribution: Subtasks inherit parent context
• Agent Mapping: Automatic agent assignment per subtask

Simplified Task Management

• JSON Task Hierarchy: Creates hierarchical JSON subtasks (impl-N.M.P)
• Context Distribution: Subtasks inherit parent context
• Basic Status Tracking: Updates task relationships only
• No Complex Synchronization: Simple parent-child relationships

Breakdown Rules

• Only pending tasks can be broken down
• Parent becomes container (not directly executable)
• Subtasks use hierarchical format: impl-N.M.P (e.g., impl-1.1.2)
• Maximum depth: 3 levels (impl-N.M.P)
• Parent-child relationships tracked in JSON only

Usage

Basic Breakdown

/task:breakdown IMPL-1

Interactive prompt:

Task: Build authentication module

Suggested subtasks:
1. Design authentication schema
2. Implement login endpoint  
3. Add JWT token handling
4. Write unit tests

Accept task breakdown? (y/n/edit): y

Auto Strategy

/task:breakdown impl-1 --strategy=auto

Automatic generation:

✅ Task impl-1 broken down:
├── impl-1.1: Design authentication schema
├── impl-1.2: Implement core auth logic
├── impl-1.3: Add security middleware
└── impl-1.4: Write comprehensive tests

Agents assigned:
- impl-1.1 → planning-agent
- impl-1.2 → code-developer
- impl-1.3 → code-developer
- impl-1.4 → test-agent

JSON files created:
- .task/impl-1.1.json
- .task/impl-1.2.json
- .task/impl-1.3.json
- .task/impl-1.4.json

Decomposition Patterns

Feature Task Pattern

Feature: "Implement shopping cart"
├── Design data model
├── Build API endpoints
├── Add state management
├── Create UI components
└── Write tests

Bug Fix Pattern

Bug: "Fix performance issue"
├── Profile and identify bottleneck
├── Implement optimization
├── Verify fix
└── Add regression test

Refactor Pattern

Refactor: "Modernize auth system"
├── Analyze current implementation
├── Design new architecture
├── Migrate incrementally
├── Update documentation
└── Deprecate old code

Context Distribution

Parent context is intelligently distributed:

{
  "parent": {
    "id": "impl-1",
    "context": {
      "requirements": ["JWT auth", "2FA support"],
      "scope": ["src/auth/*"],
      "acceptance": ["Authentication system works"],
      "inherited_from": "WFS-user-auth"
    }
  },
  "subtasks": [
    {
      "id": "impl-1.1",
      "title": "Design authentication schema",
      "status": "pending",
      "agent": "planning-agent",
      "context": {
        "requirements": ["JWT auth schema", "User model design"],
        "scope": ["src/auth/models/*"],
        "acceptance": ["Schema validates JWT tokens", "User model complete"],
        "inherited_from": "impl-1"
      },
      "relations": {
        "parent": "impl-1",
        "subtasks": [],
        "dependencies": []
      }
    }
  ]
}

Agent Assignment Logic

Based on subtask type:

• Design/Planning → planning-agent
• Implementation → code-developer
• Testing → test-agent
• Documentation → docs-agent
• Review → review-agent

Validation

Pre-breakdown Checks

1. Task exists and is valid
2. Task status is pending
3. Not already broken down
4. Workflow in IMPLEMENT phase

Post-breakdown Actions

1. Update parent status to container
2. Create subtask JSON files
3. Update parent task with subtask references
4. Update workflow session stats

Simple File Management

File Structure Created

.workflow/WFS-[topic-slug]/
├── workflow-session.json     # Session state
├── IMPL_PLAN.md             # Static planning document
└── .task/
    ├── impl-1.json          # Parent task (container) 
    ├── impl-1.1.json        # Subtask 1
    └── impl-1.2.json        # Subtask 2

Output Files

• JSON subtask files in .task/ directory
• Updated parent task JSON with subtask references
• Updated session stats in workflow-session.json

Examples

Simple Breakdown

/task:breakdown impl-1

Result:
impl-1: Build authentication (container)
├── impl-1.1: Design auth schema
├── impl-1.2: Implement auth logic  
├── impl-1.3: Add security middleware
└── impl-1.4: Write tests

Two-Level Breakdown

/task:breakdown impl-1 --depth=2

Result:  
impl-1: E-commerce checkout (container)
├── impl-1.1: Payment processing
│   ├── impl-1.1.1: Integrate gateway
│   └── impl-1.1.2: Handle transactions
├── impl-1.2: Order management
│   └── impl-1.2.1: Create order model
└── impl-1.3: Testing

Error Handling

# Task not found
❌ Task impl-5 not found

# Already broken down
⚠️ Task impl-1 already has subtasks

# Max depth exceeded
❌ Cannot create impl-1.2.3.4 (max 3 levels)

Related Commands

• /task:create - Create new tasks
• /task:execute - Execute subtasks
• /context - View task hierarchy