catlog22 3199b91255 refactor: remove external invocation references from ui-design-agent
Agent should focus on task execution, not who invokes it. Remove all
references to orchestrator commands and external callers.

Changes:
1. **Agent description** (line 19):
   - Before: "invoked by orchestrator commands (e.g., generate.md)"
   - After: Focus on autonomous execution

2. **Core Capabilities** (line 25):
   - Removed: "Invoked by: generate.md Phase 2"
   - Agent doesn't need to know its caller

3. **Execution Process** (line 250):
   - Before: "When invoked by orchestrator command"
   - After: "Standard execution flow for design generation tasks"

4. **Task prompt terminology** (line 259):
   - Changed: "orchestrator prompt" → "task prompt"

5. **Invocation Model → Task Responsibilities** (lines 295-305):
   - Before: "You are invoked by orchestrator commands..."
   - After: Focus on task responsibilities and expected output

6. **Execution Principles** (lines 310, 315):
   - Changed: "orchestrator command" → "task prompt"
   - Changed: "Each invocation" → "Each task"

7. **Performance Optimization → Generation Approach** (lines 342-348):
   - Removed: Two-layer architecture with orchestrator responsibility
   - Added: Single-pass assembly focus (matches current implementation)

8. **Scope & Boundaries** (line 359):
   - Simplified "NOT Your Responsibilities" to "Out of Scope"
   - Removed: workflow orchestration, task scheduling references
   - Focus on what agent does, not external workflow

9. **Path Handling** (lines 423-426):
   - Changed: "Orchestrator provides" → "Task prompts provide"
   - Changed: "Agent uses" → "Use" (more direct)

Result: Agent documentation now focuses on task execution without
references to external callers or orchestration.

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude <noreply@anthropic.com>
2025-11-08 21:12:06 +08:00

🚀 Claude Code Workflow (CCW)

Version License Platform

Languages: English | 中文


Claude Code Workflow (CCW) transforms AI development from simple prompt chaining into a robust, context-first orchestration system. It solves execution uncertainty and error accumulation through structured planning, deterministic execution, and intelligent multi-model orchestration.

🎉 Version 5.5: Interactive Command Guide & Enhanced Documentation

Core Improvements:

  • Command-Guide Skill - Interactive help system with CCW-help and CCW-issue triggers
  • Enhanced Command Descriptions - All 69 commands updated with detailed functionality descriptions
  • 5-Index Command System - Organized by category, use-case, relationships, and essentials
  • Smart Recommendations - Context-aware next-step suggestions for workflow guidance

See CHANGELOG.md for full details.

📚 New to CCW? Check out the Getting Started Guide for a beginner-friendly 5-minute tutorial!


Core Concepts

CCW is built on a set of core principles that differentiate it from traditional AI development approaches:

  • Context-First Architecture: Pre-defined context gathering eliminates execution uncertainty by ensuring agents have the correct information before implementation.
  • JSON-First State Management: Task states live in .task/IMPL-*.json files as the single source of truth, enabling programmatic orchestration without state drift.
  • Autonomous Multi-Phase Orchestration: Commands chain specialized sub-commands and agents to automate complex workflows with zero user intervention.
  • Multi-Model Strategy: Leverages the unique strengths of different AI models (Gemini for analysis, Codex for implementation) for superior results.
  • Hierarchical Memory System: A 4-layer documentation system provides context at the appropriate level of abstraction, preventing information overload.
  • Specialized Role-Based Agents: A suite of agents (@code-developer, @test-fix-agent, etc.) mirrors a real software team to handle diverse tasks.

⚙️ Installation

For detailed installation instructions, please refer to the INSTALL.md guide.

🚀 Quick One-Line Installation

Windows (PowerShell):

Invoke-Expression (Invoke-WebRequest -Uri "https://raw.githubusercontent.com/catlog22/Claude-Code-Workflow/main/install-remote.ps1" -UseBasicParsing).Content

Linux/macOS (Bash/Zsh):

bash <(curl -fsSL https://raw.githubusercontent.com/catlog22/Claude-Code-Workflow/main/install-remote.sh)

Verify Installation

After installation, open Claude Code and check if the workflow commands are available by running:

/workflow:session:list

If the slash commands (e.g., /workflow:*) are recognized, the installation was successful.


🛠️ Command Reference

CCW provides a rich set of commands for managing workflows, tasks, and interacting with AI tools. For a complete list and detailed descriptions of all available commands, please see the COMMAND_REFERENCE.md file.

For a detailed technical specification of every command, see the COMMAND_SPEC.md.


💡 Need Help? Use the Interactive Command Guide

CCW includes a built-in command-guide skill to help you discover and use commands effectively:

  • CCW-help - Get interactive help and command recommendations
  • CCW-issue - Report bugs or request features with guided templates

The command guide provides:

  • 🔍 Smart Command Search - Find commands by keyword, category, or use-case
  • 🤖 Next-Step Recommendations - Get suggestions for what to do after any command
  • 📖 Detailed Documentation - View parameters, examples, and best practices
  • 🎓 Beginner Onboarding - Learn the top 14 essential commands with a guided learning path
  • 📝 Issue Reporting - Generate standardized bug reports and feature requests

Example Usage:

User: "CCW-help"
→ Interactive menu with command search, recommendations, and documentation

User: "What's next after /workflow:plan?"
→ Recommends /workflow:execute, /workflow:action-plan-verify, with workflow patterns

User: "CCW-issue"
→ Guided template generation for bugs, features, or questions

🚀 Getting Started

The best way to get started is to follow the 5-minute tutorial in the Getting Started Guide.

Here is a quick example of a common development workflow:

  1. Create a Plan (automatically starts a session):
    /workflow:plan "Implement JWT-based user login and registration"
    
  2. Execute the Plan:
    /workflow:execute
    
  3. Check Status (optional):
    /workflow:status
    

🤝 Contributing & Support

📄 License

This project is licensed under the MIT License. See the LICENSE file for details.

Description
JSON-driven multi-agent development framework with intelligent CLI orchestration (Gemini/Qwen/Codex), context-first architecture, and automated workflow execution
Readme MIT 35 MiB
Languages
TypeScript 34.7%
HTML 25%
Python 21.6%
JavaScript 13.2%
CSS 4%
Other 1.5%