Files
Claude-Code-Workflow/WORKFLOW_DECISION_GUIDE_EN.md
Claude 383da9ebb7 docs: add CLI tools collaboration mode to Workflow Decision Guide
Add comprehensive section on multi-model CLI collaboration (Gemini/Qwen/Codex):
- Three execution modes: serial, parallel, and hybrid
- Semantic invocation vs command invocation patterns
- Integration examples with Lite and Full workflows
- Best practices for tool selection and execution strategies

Updates both Chinese and English versions with practical examples showing
how to leverage ultra-long context models (Gemini/Qwen) for analysis and
Codex for precise code implementation.
2025-11-20 09:51:50 +00:00

20 KiB
Raw Blame History

🌳 CCW Workflow Decision Guide

This guide helps you choose the right commands and workflows for the complete software development lifecycle.


📊 Full Lifecycle Command Selection Flowchart

flowchart TD
    Start([Start New Feature/Project]) --> Q1{Know what to build?}

    Q1 -->|No| Ideation[💡 Ideation Phase<br>Requirements Exploration]
    Q1 -->|Yes| Q2{Know how to build?}

    Ideation --> BrainIdea[/ /workflow:brainstorm:auto-parallel<br>Explore product direction and positioning /]
    BrainIdea --> Q2

    Q2 -->|No| Design[🏗️ Design Exploration<br>Architecture Solution Discovery]
    Q2 -->|Yes| Q3{Need UI design?}

    Design --> BrainDesign[/ /workflow:brainstorm:auto-parallel<br>Explore technical solutions and architecture /]
    BrainDesign --> Q3

    Q3 -->|Yes| UIDesign[🎨 UI Design Phase]
    Q3 -->|No| Q4{Task complexity?}

    UIDesign --> Q3a{Have reference design?}
    Q3a -->|Yes| UIImitate[/ /workflow:ui-design:imitate-auto<br>--input reference URL /]
    Q3a -->|No| UIExplore[/ /workflow:ui-design:explore-auto<br>--prompt design description /]

    UIImitate --> UISync[/ /workflow:ui-design:design-sync<br>Sync design system /]
    UIExplore --> UISync
    UISync --> Q4

    Q4 -->|Simple & Quick| LitePlan[⚡ Lightweight Planning<br>/workflow:lite-plan]
    Q4 -->|Complex & Complete| FullPlan[📋 Full Planning<br>/workflow:plan]

    LitePlan --> Q5{Need code exploration?}
    Q5 -->|Yes| LitePlanE[/ /workflow:lite-plan -e<br>task description /]
    Q5 -->|No| LitePlanNormal[/ /workflow:lite-plan<br>task description /]

    LitePlanE --> LiteConfirm[Three-Dimensional Confirmation:<br>1⃣ Task Approval<br>2⃣ Execution Method<br>3⃣ Code Review]
    LitePlanNormal --> LiteConfirm

    LiteConfirm --> Q6{Choose execution method}
    Q6 -->|Agent| LiteAgent[/ /workflow:lite-execute<br>Using @code-developer /]
    Q6 -->|CLI Tools| LiteCLI[CLI Execution<br>Gemini/Qwen/Codex]
    Q6 -->|Plan Only| UserImpl[Manual User Implementation]

    FullPlan --> PlanVerify{Verify plan quality?}
    PlanVerify -->|Yes| Verify[/ /workflow:action-plan-verify /]
    PlanVerify -->|No| Execute
    Verify --> Q7{Verification passed?}
    Q7 -->|No| FixPlan[Fix plan issues]
    Q7 -->|Yes| Execute
    FixPlan --> Execute

    Execute[🚀 Execution Phase<br>/workflow:execute]
    LiteAgent --> TestDecision
    LiteCLI --> TestDecision
    UserImpl --> TestDecision
    Execute --> TestDecision

    TestDecision{Need testing?}
    TestDecision -->|TDD Mode| TDD[/ /workflow:tdd-plan<br>Test-Driven Development /]
    TestDecision -->|Post-Implementation Testing| TestGen[/ /workflow:test-gen<br>Generate tests /]
    TestDecision -->|Existing Tests| TestCycle[/ /workflow:test-cycle-execute<br>Test-fix cycle /]
    TestDecision -->|No| Review

    TDD --> TDDExecute[/ /workflow:execute<br>Red-Green-Refactor /]
    TDDExecute --> TDDVerify[/ /workflow:tdd-verify<br>Verify TDD compliance /]
    TDDVerify --> Review

    TestGen --> TestExecute[/ /workflow:execute<br>Execute test tasks /]
    TestExecute --> TestResult{Tests passed?}
    TestResult -->|No| TestCycle
    TestResult -->|Yes| Review

    TestCycle --> TestPass{Pass rate ≥95%?}
    TestPass -->|No, continue fixing| TestCycle
    TestPass -->|Yes| Review

    Review[📝 Review Phase]
    Review --> Q8{Need specialized review?}
    Q8 -->|Security| SecurityReview[/ /workflow:review<br>--type security /]
    Q8 -->|Architecture| ArchReview[/ /workflow:review<br>--type architecture /]
    Q8 -->|Quality| QualityReview[/ /workflow:review<br>--type quality /]
    Q8 -->|Comprehensive| GeneralReview[/ /workflow:review<br>Comprehensive review /]
    Q8 -->|No| Complete

    SecurityReview --> Complete
    ArchReview --> Complete
    QualityReview --> Complete
    GeneralReview --> Complete

    Complete[✅ Completion Phase<br>/workflow:session:complete]
    Complete --> End([Project Complete])

    style Start fill:#e1f5ff
    style End fill:#c8e6c9
    style BrainIdea fill:#fff9c4
    style BrainDesign fill:#fff9c4
    style UIImitate fill:#f8bbd0
    style UIExplore fill:#f8bbd0
    style LitePlan fill:#b3e5fc
    style FullPlan fill:#b3e5fc
    style Execute fill:#c5e1a5
    style TDD fill:#ffccbc
    style TestGen fill:#ffccbc
    style TestCycle fill:#ffccbc
    style Review fill:#d1c4e9
    style Complete fill:#c8e6c9

🎯 Decision Point Explanations

1 Ideation Phase - "Know what to build?"

Situation Command Description
Uncertain about product direction /workflow:brainstorm:auto-parallel "Explore XXX domain product opportunities" Multi-role analysis with Product Manager, UX Expert, etc.
Clear feature requirements Skip to design phase Already know what functionality to build

Examples:

# Uncertain scenario: Want to build a collaboration tool, but unsure what exactly
/workflow:brainstorm:auto-parallel "Explore team collaboration tool positioning and core features" --count 5

# Certain scenario: Building a real-time document collaboration editor (requirements clear)
# Skip ideation, move to design phase

2 Design Phase - "Know how to build?"

Situation Command Description
Don't know technical approach /workflow:brainstorm:auto-parallel "Design XXX system architecture" System Architect, Security Expert analyze technical solutions
Clear implementation path Skip to planning Already know tech stack, architecture patterns

Examples:

# Don't know how: Real-time collaboration conflict resolution? Which algorithm?
/workflow:brainstorm:auto-parallel "Design conflict resolution mechanism for real-time collaborative document editing" --count 4

# Know how: Using Operational Transformation + WebSocket + Redis
# Skip design exploration, go directly to planning
/workflow:plan "Implement real-time collaborative editing using OT algorithm, WebSocket communication, Redis storage"

3 UI Design Phase - "Need UI design?"

Situation Command Description
🎨 Have reference design /workflow:ui-design:imitate-auto --input "URL" Copy from existing design
🎨 Design from scratch /workflow:ui-design:explore-auto --prompt "description" Generate multiple design variants
⏭️ Backend/No UI Skip Pure backend API, CLI tools, etc.

Examples:

# Have reference: Imitate Google Docs collaboration interface
/workflow:ui-design:imitate-auto --input "https://docs.google.com"

# No reference: Design from scratch
/workflow:ui-design:explore-auto --prompt "Modern minimalist document collaboration editing interface" --style-variants 3

# Sync design to project
/workflow:ui-design:design-sync --session WFS-xxx --selected-prototypes "v1,v2"

4 Planning Phase - Choose Workflow Type

Workflow Use Case Characteristics
/workflow:lite-plan Quick tasks, small features In-memory planning, three-dimensional confirmation, fast execution
/workflow:plan Complex projects, team collaboration Persistent plans, quality gates, complete traceability

Lite-Plan Three-Dimensional Confirmation:

  1. Task Approval: Confirm / Modify / Cancel
  2. Execution Method: Agent / Provide Plan / CLI Tools (Gemini/Qwen/Codex)
  3. Code Review: No / Claude / Gemini / Qwen / Codex

Examples:

# Simple task
/workflow:lite-plan "Add user avatar upload feature"

# Need code exploration
/workflow:lite-plan -e "Refactor authentication module to OAuth2 standard"

# Complex project
/workflow:plan "Implement complete real-time collaborative editing system"
/workflow:action-plan-verify  # Verify plan quality
/workflow:execute

5 Testing Phase - Choose Testing Strategy

Strategy Command Use Case
TDD Mode /workflow:tdd-plan Starting from scratch, test-driven development
Post-Implementation Testing /workflow:test-gen Code complete, add tests
Test Fixing /workflow:test-cycle-execute Existing tests, need to fix failures

Examples:

# TDD: Write tests first, then implement
/workflow:tdd-plan "User authentication module"
/workflow:execute  # Red-Green-Refactor cycle
/workflow:tdd-verify  # Verify TDD compliance

# Post-implementation testing: Add tests after code complete
/workflow:test-gen WFS-user-auth-implementation
/workflow:execute

# Test fixing: Existing tests with high failure rate
/workflow:test-cycle-execute --max-iterations 5
# Auto-iterate fixes until pass rate ≥95%

6 Review Phase - Choose Review Type

Type Command Focus
Security Review /workflow:review --type security SQL injection, XSS, authentication vulnerabilities
Architecture Review /workflow:review --type architecture Design patterns, coupling, scalability
Quality Review /workflow:review --type quality Code style, complexity, maintainability
Comprehensive Review /workflow:review All-around inspection

Examples:

# Security-critical system
/workflow:review --type security

# After architecture refactoring
/workflow:review --type architecture

# Daily development
/workflow:review --type quality

7 CLI Tools Collaboration Mode - Multi-Model Intelligent Coordination

This project integrates three CLI tools supporting flexible serial, parallel, and hybrid execution:

Tool Core Capabilities Context Length Use Cases
Gemini Deep analysis, architecture design, planning Ultra-long context Code understanding, execution flow tracing, technical solution evaluation
Qwen Code review, pattern recognition Ultra-long context Gemini alternative, multi-dimensional analysis
Codex Precise code writing, bug location Standard context Feature implementation, test generation, code refactoring

📋 Three Execution Modes

1. Serial Execution - Sequential dependency

Use case: Subsequent tasks depend on previous results

# Example: Analyze then implement
# Step 1: Gemini analyzes architecture
Use gemini to analyze the authentication module's architecture design, identify key components and data flow

# Step 2: Codex implements based on analysis
Have codex implement JWT authentication middleware based on the above architecture analysis

Execution flow:

Gemini analysis → Output architecture report → Codex reads report → Implement code

2. Parallel Execution - Concurrent processing

Use case: Multiple independent tasks with no dependencies

# Example: Multi-dimensional analysis
Use gemini to analyze authentication module security, focus on JWT, password storage, session management
Use qwen to analyze authentication module performance bottlenecks, identify slow queries and optimization points
Have codex generate unit tests for authentication module, covering all core features

Execution flow:

        ┌─ Gemini: Security analysis ─┐
Parallel ┼─ Qwen: Performance analysis ┼─→ Aggregate results
        └─ Codex: Test generation ────┘

3. Hybrid Execution - Combined serial and parallel

Use case: Complex tasks with both parallel and serial phases

# Example: Complete feature development
# Phase 1: Parallel analysis (independent tasks)
Use gemini to analyze existing authentication system architecture patterns
Use qwen to evaluate OAuth2 integration technical solutions

# Phase 2: Serial implementation (depends on Phase 1)
Have codex implement OAuth2 authentication flow based on above analysis

# Phase 3: Parallel optimization (independent tasks)
Use gemini to review code quality and security
Have codex generate integration tests

Execution flow:

Phase 1: Gemini analysis ──┐
         Qwen evaluation ──┼─→ Phase 2: Codex implementation ──→ Phase 3: Gemini review ──┐
                           │                                              Codex tests ───┼─→ Complete
                           └──────────────────────────────────────────────────────────────┘

🎯 Semantic Invocation vs Command Invocation

Method 1: Natural Language Semantic Invocation (Recommended)

# Users simply describe naturally, Claude Code auto-invokes tools
"Use gemini to analyze this module's dependencies"
→ Claude Code auto-generates: cd src && gemini -p "Analyze dependencies"

"Have codex implement user registration feature"
→ Claude Code auto-generates: codex -C src/auth --full-auto exec "Implement registration"

Method 2: Direct Command Invocation

# Precise invocation via Slash commands
/cli:chat --tool gemini "Explain this algorithm"
/cli:analyze --tool qwen "Analyze performance bottlenecks"
/cli:execute --tool codex "Optimize query performance"

🔄 Workflow Integration Examples

Integration with Lite Workflow:

# 1. Planning phase: Gemini analysis
/workflow:lite-plan -e "Refactor payment module"
→ Three-dimensional confirmation selects "CLI Tools execution"

# 2. Execution phase: Choose execution method
# Option A: Serial execution"Use gemini to analyze payment flow""Have codex refactor code"

# Option B: Parallel analysis + Serial implementation"Use gemini to analyze architecture" + "Use qwen to evaluate solution""Have codex refactor based on analysis results"

Integration with Full Workflow:

# 1. Planning phase
/workflow:plan "Implement distributed cache"
/workflow:action-plan-verify

# 2. Analysis phase (parallel)
Use gemini to analyze existing cache architecture
Use qwen to evaluate Redis cluster solution

# 3. Implementation phase (serial)
/workflow:execute  # Or use CLI
Have codex implement Redis cluster integration

# 4. Testing phase (parallel)
/workflow:test-gen WFS-cache
→ Internally uses gemini analysis + codex test generation

# 5. Review phase (serial)
Use gemini to review code quality
/workflow:review --type architecture

💡 Best Practices

When to use serial:

  • Implementation depends on design solution
  • Testing depends on code implementation
  • Optimization depends on performance analysis

When to use parallel:

  • Multi-dimensional analysis (security + performance + architecture)
  • Multi-module independent development
  • Simultaneous code and test generation

When to use hybrid:

  • Complex feature development (analysis → design → implementation → testing)
  • Large-scale refactoring (evaluation → planning → execution → verification)
  • Tech stack migration (research → solution → implementation → optimization)

Tool selection guidelines:

  1. Need to understand code → Gemini (preferred) or Qwen
  2. Need to write code → Codex
  3. Complex analysis → Gemini + Qwen parallel (complementary verification)
  4. Precise implementation → Codex (based on Gemini analysis)
  5. Quick prototype → Direct Codex usage

🔄 Complete Flow for Typical Scenarios

Scenario A: New Feature Development (Know How to Build)

# 1. Planning
/workflow:plan "Add JWT authentication and permission management"

# 2. Verify plan
/workflow:action-plan-verify

# 3. Execute
/workflow:execute

# 4. Testing
/workflow:test-gen WFS-jwt-auth
/workflow:execute

# 5. Review
/workflow:review --type security

# 6. Complete
/workflow:session:complete

Scenario B: New Feature Development (Don't Know How to Build)

# 1. Design exploration
/workflow:brainstorm:auto-parallel "Design distributed cache system architecture" --count 5

# 2. UI design (if needed)
/workflow:ui-design:explore-auto --prompt "Cache management dashboard interface"
/workflow:ui-design:design-sync --session WFS-xxx

# 3. Planning
/workflow:plan

# 4. Verification
/workflow:action-plan-verify

# 5. Execution
/workflow:execute

# 6. TDD testing
/workflow:tdd-plan "Cache system core modules"
/workflow:execute

# 7. Review
/workflow:review --type architecture
/workflow:review --type security

# 8. Complete
/workflow:session:complete

Scenario C: Quick Feature Development (Lite Workflow)

# 1. Lightweight planning (may need code exploration)
/workflow:lite-plan -e "Optimize database query performance"

# 2. Three-dimensional confirmation
# - Confirm task
# - Choose Agent execution
# - Choose Gemini code review

# 3. Auto-execution (called internally by /workflow:lite-execute)

# 4. Complete

Scenario D: Bug Fixing

# 1. Diagnosis
/cli:mode:bug-diagnosis --tool gemini "User login fails with token expired error"

# 2. Quick fix
/workflow:lite-plan "Fix JWT token expiration validation logic"

# 3. Test fix
/workflow:test-cycle-execute

# 4. Complete

🎓 Quick Command Reference

Choose by Knowledge Level

Your Situation Recommended Command
💭 Don't know what to build /workflow:brainstorm:auto-parallel "Explore product direction"
Know what, don't know how /workflow:brainstorm:auto-parallel "Design technical solution"
Know what and how /workflow:plan "Specific implementation description"
Simple, clear small task /workflow:lite-plan "Task description"
🐛 Bug fixing /cli:mode:bug-diagnosis + /workflow:lite-plan

Choose by Project Phase

Phase Command
📋 Requirements Analysis /workflow:brainstorm:auto-parallel
🏗️ Architecture Design /workflow:brainstorm:auto-parallel
🎨 UI Design /workflow:ui-design:explore-auto / imitate-auto
📝 Implementation Planning /workflow:plan / /workflow:lite-plan
🚀 Coding Implementation /workflow:execute / /workflow:lite-execute
🧪 Testing /workflow:tdd-plan / /workflow:test-gen
🔧 Test Fixing /workflow:test-cycle-execute
📖 Code Review /workflow:review
Project Completion /workflow:session:complete

Choose by Work Mode

Mode Workflow Use Case
🚀 Agile & Fast Lite Workflow Personal dev, rapid iteration, prototype validation
📋 Standard & Complete Full Workflow Team collaboration, enterprise projects, long-term maintenance
🧪 Quality-First TDD Workflow Core modules, critical features, high reliability requirements
🎨 Design-Driven UI-Design Workflow Frontend projects, user interfaces, design systems

💡 Expert Advice

Best Practices

  1. Use brainstorming when uncertain: Better to spend 10 minutes exploring solutions than blindly implementing and rewriting
  2. Use Full workflow for complex projects: Persistent plans facilitate team collaboration and long-term maintenance
  3. Use Lite workflow for small tasks: Complete quickly, reduce overhead
  4. Use TDD for critical modules: Test-driven development ensures quality
  5. Regularly update memory: /memory:update-related keeps context accurate

Common Pitfalls

  1. Blindly skipping brainstorming: Not exploring unfamiliar technical domains leads to rework
  2. Overusing brainstorming: Brainstorming even simple features wastes time
  3. Ignoring plan verification: Not running /workflow:action-plan-verify causes execution issues
  4. Ignoring testing: Not generating tests, code quality cannot be guaranteed
  5. Not completing sessions: Not running /workflow:session:complete causes session state confusion


Last Updated: 2025-11-20 Version: 5.8.1