diff --git a/README.md b/README.md index 8ea2b368..54581011 100644 --- a/README.md +++ b/README.md @@ -17,7 +17,7 @@
-Typing SVG +Typing SVG @@ -42,8 +42,8 @@ -### 🎯 4-Level Workflows -From `lite-lite-lite` (instant) to `brainstorm` (multi-role analysis) +### 🎯 Skill-based Workflows +From `lite-plan` (lightweight) to `brainstorm` (multi-role analysis) ### 🔄 Multi-CLI Orchestration Gemini, Qwen, Codex, Claude - auto-select or manual @@ -57,8 +57,8 @@ Background queue execution service -### 📊 Workflow Session Commands -start/resume/complete/sync sessions +### 📊 Session Lifecycle +start/resume/complete/sync workflow sessions ### 🖥 Terminal Dashboard Multi-terminal grid with execution monitor @@ -74,7 +74,7 @@ Agent-to-User interactive interface -> 📖 **New?** See [Workflow Guide](WORKFLOW_GUIDE.md) for the complete 4-level workflow system. +> 📖 **New?** See [Workflow Guide](WORKFLOW_GUIDE.md) for complete workflow documentation. --- @@ -87,37 +87,44 @@ npm install -g claude-code-workflow ccw install -m Global ``` -> **Troubleshooting**: If you see `NODE_MODULE_VERSION` mismatch errors for `better-sqlite3`, run `npm rebuild better-sqlite3`. See [FAQ - Troubleshooting](FAQ.md#better-sqlite3-node_module_version-mismatch) for details. +### Workflow Skills vs Commands -### Choose Your Workflow Level +CCW uses two types of invocations: + +| Type | Format | Examples | +|------|--------|----------| +| **Skills** | Trigger phrase (no slash) | `workflow:lite-plan`, `brainstorm`, `workflow:plan` | +| **Commands** | Slash command | `/ccw`, `/workflow/session:start`, `/issue/new` | + +### Choose Your Workflow Skill
- - - - - - - + + + + + + +
LevelCommandUse Case
2/workflow:lite-planQuick fixes, single-module features
2/workflow:lite-fixBug diagnosis and fix
2/workflow:multi-cli-planMulti-perspective analysis
3/workflow:planMulti-module development
3/workflow:tdd-planTest-driven development
4/workflow:brainstorm:auto-parallelNew features, architecture design
Skill TriggerUse Case
workflow:lite-planLightweight planning, single-module features
workflow:multi-cli-planMulti-CLI collaborative analysis
workflow:planFull planning with session persistence
workflow:tdd-planTest-driven development
workflow:test-fix-genTest generation and fix cycles
brainstormMulti-role brainstorming analysis
### Workflow Examples ```bash -# Level 2: Lightweight planning (recommended for most tasks) -/workflow:lite-plan "Add JWT authentication" -/workflow:lite-fix "User upload fails with 413 error" +# Skill triggers (no slash - just describe what you want) +workflow:lite-plan "Add JWT authentication" +workflow:plan "Implement payment gateway integration" +workflow:execute -# Level 3: Standard planning with session -/workflow:plan "Implement payment gateway integration" -/workflow:execute +# Brainstorming +brainstorm "Design real-time collaboration system" -# Level 4: Multi-role brainstorming -/workflow:brainstorm:auto-parallel "Design real-time collaboration system" --count 5 -/workflow:plan --session WFS-xxx -/workflow:execute +# Slash commands for session management +/workflow/session:start +/workflow/session:resume +/workflow/session:complete ``` --- @@ -172,49 +179,6 @@ Users can **semantically specify CLI tools** in prompts - the system automatical -
-📝 More Examples - -```text -# Single CLI invocation -User: "Use Gemini to analyze the database query performance" -→ System auto-calls: gemini CLI with analysis task - -# Collaborative analysis -User: "Use Gemini and Codex to collaboratively review the authentication flow" -→ System auto-calls: gemini + codex CLIs, synthesizes results - -# Parallel multi-perspective -User: "Have all available CLIs analyze this architecture design in parallel" -→ System auto-calls: gemini, codex, qwen in parallel → merged report - -# Sequential pipeline -User: "Use Gemini to plan the refactoring, then Codex to implement it" -→ System auto-calls: gemini (plan) → codex (implement) sequentially -``` - -
- -### Custom CLI Registration - -Register **any API as a custom CLI** via Dashboard interface: - -```bash -ccw view # Open Dashboard → Status → API Settings → Add Custom CLI -``` - -
- -| Field | Example | -|-------|---------| -| **Name** | `deepseek` | -| **Endpoint** | `https://api.deepseek.com/v1/chat` | -| **API Key** | `your-api-key` | - -
- -> ⚙ Register once, invoke semantically forever - no code changes needed. - --- ## 🔍 ACE Tool Configuration @@ -245,42 +209,24 @@ ACE (Augment Context Engine) provides powerful semantic code search. -
-📊 Installation - -```bash -# Enter codex-lens directory -cd codex-lens - -# Install dependencies -pip install -e . - -# Initialize index -codexlens index /path/to/project -``` - -Open Dashboard via `ccw view`, manage indexes and execute searches in **CodexLens Manager**. - -
- --- ## 💻 CCW CLI Commands -### 🌟 Recommended Commands (Main Features) +### 🌟 Recommended Commands
- - + + - - + +
CommandDescriptionWhen to Use
/ccwAuto workflow orchestrator - analyzes intent, selects workflow level, executes command chain in main process✅ General tasks, auto workflow selection, quick developmentAuto workflow orchestrator - analyzes intent, selects workflow, executes✅ General tasks, auto workflow selection
/ccw-coordinatorSmart orchestrator - intelligently recommends command chains, allows manual adjustment, executes via external CLI with state persistence🔧 Complex multi-step workflows, customizable chains, resumable sessionsSmart orchestrator - recommends command chains, allows manual adjustment🔧 Complex multi-step workflows
@@ -288,25 +234,33 @@ Open Dashboard via `ccw view`, manage indexes and execute searches in **CodexLen **Quick Examples**: ```bash -# /ccw - Auto workflow selection (Main Process) -/ccw "Add user authentication" # Auto-selects workflow based on intent -/ccw "Fix memory leak in WebSocket" # Detects bugfix workflow -/ccw "Implement with TDD" # Routes to TDD workflow +# /ccw - Auto workflow selection +/ccw "Add user authentication" +/ccw "Fix memory leak in WebSocket" +/ccw "Implement with TDD" -# /ccw-coordinator - Manual chain orchestration (External CLI) -/ccw-coordinator "Implement OAuth2 system" # Analyzes → Recommends chain → User confirms → Executes +# /ccw-coordinator - Manual chain orchestration +/ccw-coordinator "Implement OAuth2 system" ``` -**Key Differences**: +### Session Management Commands -| Aspect | /ccw | /ccw-coordinator | -|--------|------|------------------| -| **Execution** | Main process (SlashCommand) | External CLI (background tasks) | -| **Selection** | Auto intent-based | Smart recommendation + optional adjustment | -| **State** | TodoWrite tracking | Persistent state.json | -| **Use Case** | General tasks, quick dev | Complex chains, resumable | +```bash +/workflow:session:start # Start new workflow session +/workflow:session:resume # Resume paused session +/workflow:session:list # List all sessions +/workflow:session:sync # Sync session work +/workflow:session:complete # Complete session +``` ---- +### Issue Workflow Commands + +```bash +/issue/new # Create new issue +/issue/plan # Plan issue resolution +/issue/queue # Form execution queue +/issue/execute # Execute issue queue +``` ### Other CLI Commands @@ -317,45 +271,17 @@ ccw cli -p "..." # Execute CLI tools (Gemini/Qwen/Codex) ccw upgrade -a # Upgrade all installations ``` -### Dashboard Features - -
- - - - - - -
FeatureDescription
Session OverviewTrack workflow sessions and progress
CodexLensFTS + Semantic + Hybrid code search
Graph ExplorerInteractive code relationship visualization
CLI ManagerExecution history with session resume
-
- ---- - -## 📖 Documentation - -
- -| Document | Description | -|----------|-------------| -| [**Workflow Guide**](WORKFLOW_GUIDE.md) | 4-level workflow system (recommended) | -| [**Getting Started**](GETTING_STARTED.md) | 5-minute quick start | -| [**Dashboard Guide**](DASHBOARD_GUIDE.md) | Dashboard user guide | -| [**FAQ**](FAQ.md) | Common questions | -| [**Changelog**](CHANGELOG.md) | Version history | - -
- --- ## 🏗 Architecture ``` ┌─────────────────────────────────────────────────────────────────┐ -│ Main Workflow (4 Levels) │ -│ ⚡ Level 1: lite-lite-lite (instant execution) │ -│ 📝 Level 2: lite-plan / lite-fix / multi-cli-plan (→ execute) │ -│ 📊 Level 3: plan / tdd-plan / test-fix-gen (session persist) │ -│ 🧠 Level 4: brainstorm:auto-parallel → plan → execute │ +│ Workflow Skills │ +│ 📝 workflow:lite-plan / workflow:multi-cli-plan (lightweight) │ +│ 📊 workflow:plan / workflow:tdd-plan (session-based) │ +│ 🧪 workflow:test-fix-gen / workflow:test-cycle-execute │ +│ 🧠 brainstorm (multi-role analysis) │ └─────────────────────────────────────────────────────────────────┘ ↓ ┌─────────────────────────────────────────────────────────────────┐ @@ -380,13 +306,6 @@ ccw upgrade -a # Upgrade all installations └─────────────────────────────────────────────────────────────────┘ ``` -**Core Principles:** -- ⚡ **Dependency Analysis** solves parallelism - no worktree needed for main workflow -- 🀖 **Team Architecture v2** provides unified role-based agent execution with inner loop -- 🔧 **Queue Scheduler** handles background task execution with unified settings -- 🖥 **Terminal Dashboard** provides real-time monitoring and control -- 🎯 Select workflow level based on complexity - avoid over-engineering - --- ## 📊 Project Structure @@ -396,27 +315,26 @@ Claude-Code-Workflow/ ├── .claude/ │ ├── agents/ # 22 specialized agents (team-worker, cli-discuss, etc.) │ ├── commands/ # Slash commands (5 categories) -│ │ ├── root/ # Root commands (/ccw, /ccw-coordinator) +│ │ ├── ccw.md # Main orchestrator +│ │ ├── ccw-coordinator.md │ │ ├── cli/ # CLI commands (cli-init, codex-review) │ │ ├── issue/ # Issue management (plan, execute, queue) │ │ ├── memory/ # Memory commands (prepare, style-skill-memory) -│ │ └── workflow/ # Workflow commands (brainstorm, plan, session) +│ │ └── workflow/ # Workflow commands (session, ui-design, etc.) │ └── skills/ # 37 modular skills -│ ├── review-code/ # Code review with rule-based analysis -│ ├── skill-tuning/ # Skill diagnosis and optimization -│ ├── skill-generator/ # Skill scaffolding and generation -│ ├── spec-generator/ # Product specification generation -│ ├── memory-*/ # Memory management skills -│ ├── workflow-*/ # Workflow orchestration skills -│ ├── team-*/ # Team coordination skills +│ ├── workflow-lite-plan/ +│ ├── workflow-plan/ +│ ├── workflow-tdd/ +│ ├── workflow-test-fix/ +│ ├── brainstorm/ +│ ├── team-*/ # Team coordination skills │ └── ... ├── ccw/ │ ├── src/ # TypeScript source code │ │ ├── commands/ # CLI command implementations │ │ ├── core/ # Core services (a2ui, auth, hooks, routes) │ │ ├── mcp-server/ # MCP server implementation -│ │ ├── tools/ # Tool implementations -│ │ └── utils/ # Utility functions +│ │ └── tools/ # Tool implementations │ └── frontend/ # React frontend (Terminal Dashboard, Orchestrator) ├── codex-lens/ # Local semantic code search engine └── docs/ # Documentation @@ -445,13 +363,6 @@ Beat Cycle (single beat) callback <-----------------------------------------+ (next beat) SendMessage + TaskUpdate(completed) ====================================================================== - - Fast-Advance (skips coordinator for simple linear successors) -====================================================================== - [Worker A] Phase 5 complete - +- 1 ready task? simple successor? --> spawn team-worker B directly - +- complex case? --> SendMessage to coordinator -====================================================================== ``` **Key Benefits:** @@ -485,10 +396,6 @@ Visual workflow template editor with drag-drop. - 🔧 Property panel for node configuration - ⚡ Slash command integration -### Analysis Viewer - -Grid layout for analysis sessions with filtering and fullscreen mode. - --- ## 🀝 Contributing @@ -496,7 +403,6 @@ Grid layout for analysis sessions with filtering and fullscreen mode.
GitHub Issues - Contributing
--- diff --git a/README_CN.md b/README_CN.md index 1bedf27e..900eac0d 100644 --- a/README_CN.md +++ b/README_CN.md @@ -17,7 +17,7 @@
-Typing SVG +Typing SVG @@ -42,8 +42,8 @@ -### 🎯 4 级工䜜流 -从 `lite-lite-lite`即时执行到 `brainstorm`倚角色分析 +### 🎯 Skill 工䜜流 +从 `lite-plan`蜻量规划到 `brainstorm`倚角色分析 ### 🔄 倚 CLI 猖排 Gemini、Qwen、Codex、Claude - 自劚选择或手劚指定 @@ -57,8 +57,8 @@ Gemini、Qwen、Codex、Claude - 自劚选择或手劚指定 -### 📊 工䜜流䌚话呜什 -启劚/恢倍/完成/同步䌚话 +### 📊 䌚话生呜呚期 +启劚/恢倍/完成/同步工䜜流䌚话 ### 🖥 终端仪衚板 倚终端眑栌垊执行监控噚 @@ -74,7 +74,7 @@ Gemini、Qwen、Codex、Claude - 自劚选择或手劚指定 -> 📖 **新甚户** 查看 [工䜜流指南](WORKFLOW_GUIDE_CN.md) 了解完敎的 4 级工䜜流系统。 +> 📖 **新甚户** 查看 [工䜜流指南](WORKFLOW_GUIDE_CN.md) 了解完敎的工䜜流文档。 --- @@ -87,35 +87,44 @@ npm install -g claude-code-workflow ccw install -m Global ``` -### 选择工䜜流级别 +### 工䜜流 Skill 䞎呜什 + +CCW 䜿甚䞀种调甚方匏 + +| 类型 | 栌匏 | 瀺䟋 | +|------|------|------| +| **Skills** | 觊发短语无斜杠 | `workflow:lite-plan`, `brainstorm`, `workflow:plan` | +| **Commands** | 斜杠呜什 | `/ccw`, `/workflow/session:start`, `/issue/new` | + +### 选择工䜜流 Skill
- - - - - - - + + + + + + +
级别呜什䜿甚场景
2/workflow:lite-plan快速修倍、单暡块功胜
2/workflow:lite-fixBug 诊断修倍
2/workflow:multi-cli-plan倚视角分析
3/workflow:plan倚暡块匀发
3/workflow:tdd-plan测试驱劚匀发
4/workflow:brainstorm:auto-parallel新功胜、架构讟计
Skill 觊发词䜿甚场景
workflow:lite-plan蜻量规划、单暡块功胜
workflow:multi-cli-plan倚 CLI 协同分析
workflow:plan完敎规划䞎䌚话持久化
workflow:tdd-plan测试驱劚匀发
workflow:test-fix-gen测试生成䞎修倍埪环
brainstorm倚角色倎脑风暎分析
### 工䜜流瀺䟋 ```bash -# Level 2: 蜻量规划 (掚荐甚于倧倚数任务) -/workflow:lite-plan "添加 JWT 讀证" -/workflow:lite-fix "甚户䞊䌠倱莥返回 413 错误" +# Skill 觊发无斜杠 - 盎接描述䜠想做什么 +workflow:lite-plan "添加 JWT 讀证" +workflow:plan "实现支付眑关集成" +workflow:execute -# Level 3: 标准规划 + Session -/workflow:plan "实现支付眑关集成" -/workflow:execute +# 倎脑风暎 +brainstorm "讟计实时协䜜系统" -# Level 4: 倚角色倎脑风暎 -/workflow:brainstorm:auto-parallel "讟计实时协䜜系统" --count 5 -/workflow:plan --session WFS-xxx -/workflow:execute +# 䌚话管理呜什 +/workflow:session:start +/workflow:session:resume +/workflow:session:complete ``` --- @@ -170,49 +179,6 @@ ccw install -m Global -
-📝 曎倚瀺䟋 - -```text -# 单 CLI 调甚 -甚户: "䜿甚 Gemini 分析数据库查询性胜" -→ 系统自劚调甚: gemini CLI 执行分析任务 - -# 协同分析 -甚户: "䜿甚 Gemini 和 Codex 协同审查讀证流皋" -→ 系统自劚调甚: gemini + codex CLI绌合分析结果 - -# 并行倚视角 -甚户: "让所有可甚的 CLI 并行分析这䞪架构讟计" -→ 系统自劚调甚: gemini, codex, qwen 并行执行 → 合并报告 - -# 顺序流氎线 -甚户: "甹 Gemini 规划重构方案然后 Codex 实现" -→ 系统自劚调甚: gemini规划→ codex实现顺序执行 -``` - -
- -### 自定义 CLI 泚册 - -通过 Dashboard 界面 **泚册任意 API 䞺自定义 CLI** - -```bash -ccw view # 打匀 Dashboard → Status → API Settings → 添加自定义 CLI -``` - -
- -| 字段 | 瀺䟋 | -|------|------| -| **名称** | `deepseek` | -| **端点** | `https://api.deepseek.com/v1/chat` | -| **API Key** | `your-api-key` | - -
- -> ⚙ 泚册䞀次氞久语义调甚 - 无需修改代码。 - --- ## 🔍 ACE Tool 配眮 @@ -243,42 +209,24 @@ ACE (Augment Context Engine) 提䟛区倧的语义代码搜玢胜力。 -
-📊 安装 - -```bash -# 进入 codex-lens 目圕 -cd codex-lens - -# 安装䟝赖 -pip install -e . - -# 初始化玢匕 -codexlens index /path/to/project -``` - -通过 `ccw view` 打匀 Dashboard圚 **CodexLens Manager** 䞭管理玢匕和执行搜玢。 - -
- --- ## 💻 CCW CLI 呜什 -### 🌟 掚荐呜什栞心功胜 +### 🌟 掚荐呜什
- - + + - - + +
呜什诎明适甚场景
/ccw自劚工䜜流猖排噚 - 分析意囟、自劚选择工䜜流级别、圚䞻进皋䞭执行呜什铟✅ 通甚任务、自劚选择工䜜流、快速匀发自劚工䜜流猖排噚 - 分析意囟、选择工䜜流、执行✅ 通甚任务、自劚选择工䜜流
/ccw-coordinator智胜猖排噚 - 智胜掚荐呜什铟、支持手劚调敎、通过倖郚 CLI 执行、持久化状态🔧 倍杂倚步骀工䜜流、可自定义铟、可恢倍䌚话智胜猖排噚 - 掚荐呜什铟、支持手劚调敎🔧 倍杂倚步骀工䜜流
@@ -286,25 +234,33 @@ codexlens index /path/to/project **快速瀺䟋** ```bash -# /ccw - 自劚工䜜流选择䞻进皋 -/ccw "添加甚户讀证" # 自劚根据意囟选择工䜜流 -/ccw "修倍 WebSocket 䞭的内存泄挏" # 识别䞺 bugfix 工䜜流 -/ccw "䜿甚 TDD 方匏实现" # 路由到 TDD 工䜜流 +# /ccw - 自劚工䜜流选择 +/ccw "添加甚户讀证" +/ccw "修倍 WebSocket 䞭的内存泄挏" +/ccw "䜿甚 TDD 方匏实现" -# /ccw-coordinator - 手劚铟猖排倖郚 CLI -/ccw-coordinator "实现 OAuth2 系统" # 分析 → 掚荐铟 → 甚户确讀 → 执行 +# /ccw-coordinator - 手劚铟猖排 +/ccw-coordinator "实现 OAuth2 系统" ``` -**䞻芁区别** +### 䌚话管理呜什 -| 方面 | /ccw | /ccw-coordinator | -|------|------|------------------| -| **执行方匏** | 䞻进皋SlashCommand | 倖郚 CLI后台任务 | -| **选择方匏** | 自劚基于意囟识别 | 智胜掚荐 + 可选调敎 | -| **状态管理** | TodoWrite 跟螪 | 持久化 state.json | -| **适甚场景** | 通甚任务、快速匀发 | 倍杂铟条、可恢倍 | +```bash +/workflow:session:start # 启劚新工䜜流䌚话 +/workflow:session:resume # 恢倍暂停的䌚话 +/workflow:session:list # 列出所有䌚话 +/workflow:session:sync # 同步䌚话工䜜 +/workflow:session:complete # 完成䌚话 +``` ---- +### Issue 工䜜流呜什 + +```bash +/issue/new # 创建新 issue +/issue/plan # 规划 issue 解决方案 +/issue/queue # 圢成执行队列 +/issue/execute # 执行 issue 队列 +``` ### 其他 CLI 呜什 @@ -315,45 +271,17 @@ ccw cli -p "..." # 执行 CLI 工具 (Gemini/Qwen/Codex) ccw upgrade -a # 升级所有安装 ``` -### Dashboard 功胜 - -
- - - - - - -
功胜诎明
䌚话抂览跟螪工䜜流䌚话和进床
CodexLensFTS + 语义 + 混合代码搜玢
囟浏览噚亀互匏代码关系可视化
CLI 管理噚执行历史䞎䌚话恢倍
-
- ---- - -## 📖 文档 - -
- -| 文档 | 诎明 | -|------|------| -| [**工䜜流指南**](WORKFLOW_GUIDE_CN.md) | 4 级工䜜流系统掚荐 | -| [**快速匀始**](GETTING_STARTED_CN.md) | 5 分钟快速入闚 | -| [**Dashboard 指南**](DASHBOARD_GUIDE.md) | Dashboard 甚户指南 | -| [**垞见问题**](FAQ.md) | 垞见问题解答 | -| [**曎新日志**](CHANGELOG.md) | 版本历史 | - -
- --- ## 🏗 架构抂览 ``` ┌─────────────────────────────────────────────────────────────────┐ -│ 䞻干工䜜流 (4 级) │ -│ ⚡ Level 1: lite-lite-lite (即时执行) │ -│ 📝 Level 2: lite-plan / lite-fix / multi-cli-plan (→ execute) │ -│ 📊 Level 3: plan / tdd-plan / test-fix-gen (Session 持久化) │ -│ 🧠 Level 4: brainstorm:auto-parallel → plan → execute │ +│ 工䜜流 Skills │ +│ 📝 workflow:lite-plan / workflow:multi-cli-plan (蜻量级) │ +│ 📊 workflow:plan / workflow:tdd-plan (䌚话匏) │ +│ 🧪 workflow:test-fix-gen / workflow:test-cycle-execute │ +│ 🧠 brainstorm (倚角色分析) │ └─────────────────────────────────────────────────────────────────┘ ↓ ┌─────────────────────────────────────────────────────────────────┐ @@ -378,13 +306,6 @@ ccw upgrade -a # 升级所有安装 └─────────────────────────────────────────────────────────────────┘ ``` -**栞心原则** -- ⚡ **䟝赖分析** 解决并行问题 - 䞻干工䜜流无需 worktree -- 🀖 **Team 架构 v2** 提䟛统䞀的基于角色的智胜䜓执行䞎内埪环 -- 🔧 **队列调床噚** 倄理后台任务执行䞎统䞀讟眮 -- 🖥 **终端仪衚板** 提䟛实时监控和控制 -- 🎯 根据倍杂床选择工䜜流级别 - 避免过床工皋化 - --- ## 📊 项目结构 @@ -394,27 +315,26 @@ Claude-Code-Workflow/ ├── .claude/ │ ├── agents/ # 22 䞪䞓䞚化智胜䜓 (team-worker, cli-discuss 等) │ ├── commands/ # 斜杠呜什5 䞪类别 -│ │ ├── root/ # 根呜什 (/ccw, /ccw-coordinator) +│ │ ├── ccw.md # 䞻猖排噚 +│ │ ├── ccw-coordinator.md │ │ ├── cli/ # CLI 呜什 (cli-init, codex-review) │ │ ├── issue/ # Issue 管理 (plan, execute, queue) │ │ ├── memory/ # 内存呜什 (prepare, style-skill-memory) -│ │ └── workflow/ # 工䜜流呜什 (brainstorm, plan, session) +│ │ └── workflow/ # 工䜜流呜什 (session, ui-design 等) │ └── skills/ # 37 䞪暡块化技胜 -│ ├── review-code/ # 基于规则的代码审查 -│ ├── skill-tuning/ # 技胜诊断䞎䌘化 -│ ├── skill-generator/ # 技胜脚手架䞎生成 -│ ├── spec-generator/ # 产品规栌生成 -│ ├── memory-*/ # 内存管理技胜 -│ ├── workflow-*/ # 工䜜流猖排技胜 -│ ├── team-*/ # 团队协调技胜 +│ ├── workflow-lite-plan/ +│ ├── workflow-plan/ +│ ├── workflow-tdd/ +│ ├── workflow-test-fix/ +│ ├── brainstorm/ +│ ├── team-*/ # 团队协调技胜 │ └── ... ├── ccw/ │ ├── src/ # TypeScript 源码 │ │ ├── commands/ # CLI 呜什实现 │ │ ├── core/ # 栞心服务 (a2ui, auth, hooks, routes) │ │ ├── mcp-server/ # MCP 服务噚实现 -│ │ ├── tools/ # 工具实现 -│ │ └── utils/ # 工具凜数 +│ │ └── tools/ # 工具实现 │ └── frontend/ # React 前端终端仪衚板、猖排噚 ├── codex-lens/ # 本地语义代码搜玢匕擎 └── docs/ # 文档 @@ -443,13 +363,6 @@ v2 团队架构匕入了**事件驱劚的节拍暡型**实现高效猖排 回调 <--------------------------------------+ (䞋䞀节拍) SendMessage + TaskUpdate(completed) ====================================================================== - - 快速掚进 (简单线性后继跳过协调噚) -====================================================================== - [工䜜者 A] 阶段 5 完成 - +- 1䞪就绪任务简单后继 --> 盎接生成 team-worker B - +- 倍杂情况 --> SendMessage 给协调噚 -====================================================================== ``` **栞心䌘势** @@ -483,10 +396,6 @@ v2 团队架构匕入了**事件驱劚的节拍暡型**实现高效猖排 - 🔧 节点配眮属性面板 - ⚡ 斜杠呜什集成 -### 分析查看噚 (Analysis Viewer) - -垊过滀功胜的分析䌚话眑栌垃局支持党屏暡匏。 - --- ## 🀝 莡献 @@ -494,7 +403,6 @@ v2 团队架构匕入了**事件驱劚的节拍暡型**实现高效猖排
GitHub Issues - Contributing
--- diff --git a/WORKFLOW_GUIDE.md b/WORKFLOW_GUIDE.md index 7e400c7b..87a25123 100644 --- a/WORKFLOW_GUIDE.md +++ b/WORKFLOW_GUIDE.md @@ -2,32 +2,7 @@ ## Overview -CCW provides a comprehensive workflow system built on **Team Architecture v2** and a **4-Level Workflow System**, designed to cover the complete software development lifecycle from rapid prototyping to full team orchestration. - -``` -+---------------------------------------------------------------+ -| 4-Level Workflow System | -| | -| +----------+ +----------+ +----------+ +--------------+ | -| | Level 1 | ->| Level 2 | ->| Level 3 | ->| Level 4 | | -| | Rapid | |Lightweight| | Standard | | Brainstorm | | -| | | | | | | | | | -| | lite- | |lite-plan | | plan | | brainstorm | | -| | lite-lite| |lite-fix | | tdd-plan | | :auto- | | -| | | |multi-cli-| |test-fix- | | parallel | | -| | | | plan | | gen | | | | | -| +----------+ +----------+ +----------+ +-----|--------+ | -| | | -| +------------v------+ | -| | Team Architecture | | -| | v2 | | -| | team-coordinate | | -| | team-executor | | -| +-------------------+ | -| | -| Manual: High -------------------------------------> Low: Auto | -+---------------------------------------------------------------+ -``` +CCW provides a comprehensive workflow system built on **Team Architecture v2** and **Skill-based Workflows**, designed to cover the complete software development lifecycle. ## What's New in v7.0 @@ -37,605 +12,217 @@ CCW provides a comprehensive workflow system built on **Team Architecture v2** a - **Queue Scheduler**: Background task execution with dependency resolution - **Workflow Session Commands**: `start`, `resume`, `complete`, `sync` for full lifecycle management - **Beat/Cadence Orchestration**: Event-driven coordination model -- **New Dashboard Views**: Analysis Viewer, Terminal Dashboard, Orchestrator Template Editor --- -## Table of Contents +## Skills vs Commands -1. [4-Level Workflow System](#4-level-workflow-system) -2. [Session Lifecycle](#session-lifecycle) -3. [Team Architecture v2](#team-architecture-v2) -4. [Command Categories](#command-categories) -5. [Skill Categories](#skill-categories) -6. [Workflow Selection Guide](#workflow-selection-guide) -7. [Issue Workflow](#issue-workflow) +CCW uses two types of invocations: + +| Type | Format | Examples | +|------|--------|----------| +| **Skills** | Trigger phrase (no slash) | `workflow:lite-plan`, `brainstorm`, `workflow:plan` | +| **Commands** | Slash command | `/ccw`, `/workflow/session:start`, `/issue/new` | --- -## 4-Level Workflow System +## Workflow Skills -### Level 1: Rapid Execution (lite-lite-lite) +### Lightweight Planning -**Fastest - Direct execution, minimal overhead** - -``` -+-----------------+ +------------------+ -| lite-lite-lite | --> | Direct Execution | -+-----------------+ +------------------+ -``` - -**Characteristics**: -| Property | Value | -|----------|-------| -| **Complexity** | Lowest | -| **Planning** | In-memory, immediate | -| **Artifacts** | None (code only) | -| **Use Case** | Quick fixes, config changes | - -**Commands**: -```bash -/workflow:lite-lite-lite "Add CORS headers to API" -``` - ---- - -### Level 2: Lightweight Planning - -**Fast - In-memory planning or single analysis, quick iteration** - -| Workflow | Purpose | Artifacts | Execution | -|----------|---------|-----------|-----------| -| `lite-plan` | Clear requirement development | memory://plan | -> `lite-execute` | -| `lite-fix` | Bug diagnosis and fix | `.workflow/.lite-fix/` | -> `lite-execute` | -| `multi-cli-plan` | Multi-perspective tasks | `.workflow/.multi-cli-plan/` | -> `lite-execute` | - -#### lite-plan -> lite-execute +| Skill Trigger | Purpose | Phases | +|---------------|---------|--------| +| `workflow:lite-plan` | Lightweight planning with exploration | 5 phases | +| `workflow:lite-execute` | Execute lite-plan output | Execution | **5-Phase Interactive Workflow**: ``` Phase 1: Task Analysis & Exploration (30-90s) - └─ Auto-detect codebase context need, launch @cli-explore-agent - Phase 2: Clarification (user-dependent) - └─ Interactive Q&A based on exploration findings - Phase 3: Planning (20-60s) - ├─ Low complexity -> Direct planning - └─ Medium/High complexity -> @cli-planning-agent - Phase 4: Three-Dimensional Confirmation - └─ Task approval + Execution method + Code review tool - -Phase 5: Execution & Tracking (5-120min) - └─ Live progress with selected method +Phase 5: Execution & Tracking ``` -**Key Features**: -- Smart code exploration (auto-detect or force with `-e`) -- In-memory planning (no file artifacts) -- Three-dimensional multi-select confirmation -- Flexible execution (Agent or CLI) -- Optional post-review +### Multi-CLI Planning -```bash -/workflow:lite-plan "Add JWT authentication" # Basic usage -/workflow:lite-plan -e "Refactor logging" # Force exploration -/workflow:lite-execute # Execute the plan +| Skill Trigger | Purpose | +|---------------|---------| +| `workflow:multi-cli-plan` | Multi-CLI collaborative analysis | + +**5-Phase Workflow**: ``` - -#### lite-fix - -**Intelligent Bug Diagnosis + Fix (5 phases)**: -``` -Phase 1: Bug Analysis - ├─ Severity pre-assessment (Low/Medium/High/Critical) - └─ Parallel cli-explore-agent diagnosis (1-4 angles) - -Phase 2: Clarification (optional) - └─ Aggregate clarification needs, AskUserQuestion - -Phase 3: Fix Planning - ├─ Low/Medium -> Claude direct planning - └─ High/Critical -> cli-lite-planning-agent - -Phase 4: Confirmation - └─ User confirms execution method - -Phase 5: Execute - └─ /workflow:lite-execute --in-memory --mode bugfix -``` - -```bash -/workflow:lite-fix # Standard fix -/workflow:lite-fix --hotfix # Emergency hotfix (skip diagnosis) -``` - -**Artifacts**: `.workflow/.lite-fix/{bug-slug}-{date}/` - -#### multi-cli-plan - -**Multi-CLI Collaborative Analysis + Consensus (5 phases)**: -``` -Phase 1: Context Gathering - └─ ACE semantic search, build context package - +Phase 1: Context Gathering (ACE semantic search) Phase 2: Multi-CLI Discussion (iterative) - ├─ cli-discuss-agent: Gemini + Codex + Claude - ├─ Cross-verification, synthesize solutions - └─ Loop until convergence or max rounds - Phase 3: Present Options - └─ Display solutions with trade-offs - Phase 4: User Decision - └─ User selects solution - Phase 5: Plan Generation - └─ -> lite-execute ``` -```bash -/workflow:multi-cli-plan "Compare OAuth vs JWT" +### Standard Planning + +| Skill Trigger | Purpose | Phases | +|---------------|---------|--------| +| `workflow:plan` | Full planning with session | 5 phases | +| `workflow:plan-verify` | Plan verification | Verification | +| `workflow:replan` | Interactive replanning | Replanning | + +### TDD Workflow + +| Skill Trigger | Purpose | +|---------------|---------| +| `workflow:tdd-plan` | TDD planning | +| `workflow:tdd-verify` | TDD verification | + +**6-Phase TDD Planning + Red-Green-Refactor**: +``` +Phase 1: Test Design +Phase 2: Red (write failing test) +Phase 3: Green (minimal implementation) +Phase 4: Refactor +Phase 5: Verify +Phase 6: Next cycle ``` ---- +### Test-Fix Workflow -### Level 3: Standard Planning +| Skill Trigger | Purpose | +|---------------|---------| +| `workflow:test-fix-gen` | Test generation and fix | +| `workflow:test-cycle-execute` | Execute test cycles | -**Complete - Persistent Session + Verification + Full Execution** +**Progressive Test Layers (L0-L3)**: -| Workflow | Purpose | Phases | Artifacts | -|----------|---------|--------|-----------| -| `plan` | Complex feature development | 5 phases | `.workflow/active/{session}/` | -| `tdd-plan` | Test-driven development | 6 phases | `.workflow/active/{session}/` | -| `test-fix-gen` | Test fix generation | 5 phases | `.workflow/active/WFS-test-{session}/` | - -#### plan -> verify -> execute - -**5-Phase Complete Planning**: -``` -Phase 1: Session Discovery - └─ /workflow:session:start --auto - -Phase 2: Context Gathering - └─ /workflow:tools:context-gather - └─ Returns context-package.json + conflict_risk - -Phase 3: Conflict Resolution (conditional) - └─ IF conflict_risk >= medium -> /workflow:tools:conflict-resolution - -Phase 4: Task Generation - └─ /workflow:tools:task-generate-agent - └─ Returns IMPL_PLAN.md + IMPL-*.json + TODO_LIST.md - -Phase 5: Summary + Next Steps -``` - -```bash -/workflow:plan "Multi-module refactoring" # Complete planning -/workflow:plan-verify # Verify plan (recommended) -/workflow:execute # Execute -/workflow:review # (optional) Review -``` - -**Artifacts**: `.workflow/active/{WFS-session}/` -- `workflow-session.json` -- `IMPL_PLAN.md` -- `TODO_LIST.md` -- `.task/IMPL-*.json` - -#### tdd-plan -> execute -> tdd-verify - -**6-Phase Test-Driven Development**: -``` -Phase 1: Session Discovery (--type tdd) -Phase 2: Context Gathering -Phase 3: Test Coverage Analysis -Phase 4: Conflict Resolution (conditional) -Phase 5: TDD Task Generation (Red-Green-Refactor cycles) -Phase 6: TDD Structure Validation -``` - -```bash -/workflow:tdd-plan "User authentication" -/workflow:execute -/workflow:tdd-verify -``` - -**TDD Task Structure**: -- Each IMPL task contains complete Red-Green-Refactor internal cycle -- `meta.tdd_workflow: true` -- `flow_control.implementation_approach` has 3 steps (red/green/refactor) - -#### test-fix-gen -> test-cycle-execute - -**5-Phase Test Fix Generation**: - -**Dual-mode support**: -| Mode | Input | Context Source | -|------|-------|----------------| -| Session Mode | `WFS-xxx` | Source session summaries | -| Prompt Mode | Text/file path | Direct codebase analysis | - -```bash -/workflow:test-fix-gen WFS-user-auth-v2 # Session Mode -/workflow:test-fix-gen "Test the auth API" # Prompt Mode -/workflow:test-cycle-execute # Execute test-fix cycle -``` - ---- - -### Level 4: Brainstorming (brainstorm:auto-parallel) - -**Exploratory - Multi-role brainstorming + Complete planning** - -**3-Phase Flow**: -``` -Phase 1: Interactive Framework Generation - └─ /workflow:brainstorm:artifacts - ├─ Topic analysis, generate questions - ├─ Role selection (user confirmation) - ├─ Role question collection - └─ Generate guidance-specification.md - -Phase 2: Parallel Role Analysis - └─ N x Task(conceptual-planning-agent) - ├─ Each role analyzes independently - └─ Parallel generate {role}/analysis.md - -Phase 3: Synthesis Integration - └─ /workflow:brainstorm:synthesis - └─ Integrate all analyses -> synthesis-specification.md -``` - -```bash -/workflow:brainstorm:auto-parallel "Real-time collaboration system" -/workflow:plan --session {sessionId} -/workflow:execute -``` - -**Available Roles**: -| Role | Description | -|------|-------------| -| `system-architect` | System Architect | -| `ui-designer` | UI Designer | -| `ux-expert` | UX Expert | -| `product-manager` | Product Manager | -| `data-architect` | Data Architect | -| `test-strategist` | Test Strategist | - -**Artifact Structure**: -``` -.workflow/active/WFS-{topic}/ -├── workflow-session.json -└── .brainstorming/ - ├── guidance-specification.md - ├── {role}/analysis.md - └── synthesis-specification.md -``` +| Layer | Name | Focus | +|-------|------|-------| +| **L0** | Static Analysis | Compilation, imports, types, AI code issues | +| **L1** | Unit Tests | Function/class behavior | +| **L2** | Integration Tests | Component interactions, API contracts | +| **L3** | E2E Tests | User journeys, critical paths | --- ## Session Lifecycle -CCW v7.0 introduces comprehensive session lifecycle commands for managing workflow sessions from creation to completion. - -### Session Commands Overview - -| Command | Purpose | When to Use | -|---------|---------|-------------| -| `/workflow:session:start` | Start new or discover existing | Beginning any workflow | -| `/workflow:session:resume` | Resume a paused session | Returning to interrupted work | -| `/workflow:session:complete` | Archive session and extract learnings | After all tasks complete | -| `/workflow:session:sync` | Sync session work to specs | Update project documentation | - -### Starting a Session +### Session Commands ```bash -# Discovery mode - list active sessions -/workflow:session:start - -# Auto mode - intelligently create or reuse -/workflow:session:start --auto "Implement OAuth2" - -# Force new mode -/workflow:session:start --new "User authentication" - -# Specify session type -/workflow:session:start --type tdd --auto "Test-driven login" +/workflow:session:start # Start new workflow session +/workflow:session:resume # Resume paused session +/workflow:session:list # List all sessions +/workflow:session:sync # Sync session work +/workflow:session:complete # Complete session +/workflow:session:solidify # Crystallize learnings into permanent memory ``` -**Session Types**: -- `workflow`: Standard implementation (default) -- `review`: Code review sessions -- `tdd`: Test-driven development -- `test`: Test generation/fix sessions -- `docs`: Documentation sessions +### Session Types -### Completing a Session - -```bash -/workflow:session:complete # Interactive completion -/workflow:session:complete --yes # Auto-complete with sync -/workflow:session:complete --detailed # With metrics -``` - -**Completion Actions**: -- Archive session to `.workflow/archives/` -- Generate `manifest.json` with metrics -- Extract lessons learned -- Auto-sync project state (with `--yes`) +| Type | Prefix | Description | +|------|--------|-------------| +| **Workflow** | `WFS-` | General development sessions | +| **Review** | `WFS-review-` | Code review sessions | +| **TDD** | `WFS-tdd-` | TDD workflow sessions | +| **Test** | `WFS-test-` | Test generation sessions | ### Session Directory Structure ``` -.workflow/ -├── active/ # Active sessions -│ └── WFS-{session-name}/ -│ ├── workflow-session.json # Session metadata -│ ├── IMPL_PLAN.md # Implementation plan -│ ├── TODO_LIST.md # Task checklist -│ ├── .task/ # Task JSON files -│ └── .process/ # Process artifacts -├── archives/ # Completed sessions -└── project-tech.json # Project tech registry +.workflow/active/{session-id}/ +├── workflow-session.json # Session metadata +├── IMPL_PLAN.md # Implementation plan +├── TODO_LIST.md # Task checklist +├── .task/ # Task JSON files +└── .process/ # Process artifacts ``` --- ## Team Architecture v2 -**For complex multi-role projects requiring specialized expertise and orchestration.** +### Core Concepts -### Overview +- **team-worker agent**: Unified worker agent for all roles +- **role-spec files**: Lightweight YAML frontmatter + Phase 2-4 logic +- **Inner loop framework**: Batch processing for same-prefix tasks +- **Beat/Cadence model**: Event-driven coordination -Team Architecture v2 (`team-coordinate-v2`, `team-executor-v2`, `team-lifecycle-v5`) provides a unified team-worker agent architecture for complex software development workflows. +### Available Team Skills -``` -+---------------------------------------------------------------+ -| Team Coordinate / Team Executor v2 | -| | -| +-------------+ +-------------------------------------+ | -| | Coordinator | ---> | Dynamic Role-Spec Generation | | -| | / Executor | | (analyst, planner, executor, etc.) | | -| +-------------+ +-------------------------------------+ | -| | | | -| v v | -| +-------------+ +-------------------------------------+ | -| | Task | | team-worker Agents | | -| | Dispatching | | Phase 1: Task Discovery (built-in) | | -| +-------------+ | Phase 2-4: Role-Specific (spec) | | -| | Phase 5: Report (built-in) | | -| +-------------------------------------+ | -| | | -| v | -| +-------------------------------------+ | -| | Subagents (Discuss, Explore, Docs) | | -| +-------------------------------------+ | -+---------------------------------------------------------------+ -``` - -### team-worker Agent - -The unified worker agent: -- **Phase 1 (Built-in)**: Task discovery - filters tasks by prefix and status -- **Phase 2-4 (Role-Specific)**: Loads domain logic from role-spec markdown files -- **Phase 5 (Built-in)**: Report + Fast-Advance - handles completion and successor spawning - -### Role-Spec Files - -Lightweight markdown files with YAML frontmatter: - -```yaml ---- -role: analyst -prefix: RESEARCH -inner_loop: false -subagents: [explore, discuss] -message_types: - success: research_ready - error: error ---- -``` +| Skill | Purpose | +|-------|---------| +| `team-coordinate-v2` | Dynamic role generation and coordination | +| `team-executor-v2` | Pure execution of existing sessions | +| `team-lifecycle-v5` | Full lifecycle (spec -> impl -> test) | +| `team-brainstorm` | Brainstorming team | +| `team-frontend` | Frontend development team | +| `team-testing` | Testing team | +| `team-review` | Code review team | ### Available Roles -| Role | Prefix | Responsibility | Inner Loop | -|------|--------|----------------|------------| -| analyst | RESEARCH | Codebase exploration, analysis | No | -| writer | DRAFT | Document generation | Yes | -| planner | PLAN | Task breakdown, dependency planning | Yes | -| executor | IMPL | Implementation and coding | Yes | -| tester | TEST | Testing and quality assurance | Yes | -| reviewer | REVIEW | Code review and quality gates | No | -| architect | ARCH | Architecture decisions | No | -| fe-developer | FE-IMPL | Frontend implementation | No | -| fe-qa | FE-TEST | Frontend testing | No | - -### Inner Loop Framework - -When `inner_loop: true`, a single agent processes all same-prefix tasks sequentially: - -``` -context_accumulator = [] - -Phase 1: Find first IMPL-* task - Phase 2-4: Execute role spec - Phase 5-L: Mark done, log, accumulate - More IMPL-* tasks? -> Phase 1 (loop) - No more? -> Phase 5-F (final report) -``` - -### Beat/Cadence Orchestration - -**Event-driven coordination model**: - -``` -Beat Cycle (single beat) -====================================================================== - Event Coordinator Workers ----------------------------------------------------------------------- - callback/resume --> +- handleCallback -+ - | mark completed | - | check pipeline | - +- handleSpawnNext -+ - | find ready tasks | - | spawn workers ---+--> [team-worker A] - | (parallel OK) --+--> [team-worker B] - +- STOP (idle) -----+ | - | - callback <-----------------------------------------+ - (next beat) -====================================================================== -``` - -### Commands - -#### Team Coordinate - -Generate role-specs and orchestrate a team from scratch: - -```bash -/team-coordinate "Design and implement real-time collaboration" -``` - -**Process**: -1. Analyze requirements and detect capabilities -2. Generate role-specs dynamically -3. Create tasks with dependency chains -4. Spawn team-worker agents -5. Monitor progress via callbacks -6. Complete with comprehensive report - -#### Team Executor - -Execute a pre-planned team session: - -```bash -/team-executor # Initial execution -/team-executor resume # Resume paused session -/team-executor status # Check status -``` - -#### Team Lifecycle v5 - -Unified team skill for full lifecycle (spec -> impl -> test -> review): - -```bash -# Triggers on "team lifecycle" -/team-lifecycle "Build a user authentication system with OAuth2" -``` - -**Pipeline Definitions**: - -| Pipeline | Tasks | Flow | -|----------|-------|------| -| Spec-only | 6 | RESEARCH -> DRAFT-001..004 -> QUALITY | -| Impl-only | 4 | PLAN -> IMPL -> TEST + REVIEW | -| Full-lifecycle | 10 | [Spec] -> PLAN -> IMPL -> TEST + REVIEW | -| Frontend | 3+ | PLAN -> DEV-FE -> QA-FE (GC loop) | - -### Subagents - -| Subagent | Purpose | -|----------|---------| -| discuss | Multi-perspective critique with dynamic perspectives | -| explore | Codebase exploration with caching | -| doc-generation | Document generation from templates | - -### Message Bus Protocol - -```javascript -mcp__ccw-tools__team_msg({ - operation: "log", - team: "", // Session ID, NOT team name - from: "", - to: "coordinator", - type: "", - summary: "[] ", - ref: "" -}) -``` - -### Session Structure - -``` -.workflow/.team// -├── team-session.json # Session metadata -├── task-analysis.json # Task dependencies -├── role-specs/ # Generated role-spec files -├── artifacts/ # Task outputs -├── discussions/ # Multi-perspective critiques -├── explorations/ # Cached exploration results -└── wisdom/ # Accumulated learnings - ├── learnings.md - ├── decisions.md - ├── conventions.md - └── issues.md -``` +| Role | Responsibility | +|------|----------------| +| analyst | Code analysis, requirements | +| writer | Documentation, content | +| planner | Planning, architecture | +| executor | Implementation, coding | +| tester | Testing, QA | +| reviewer | Code review, feedback | +| architect | System design, architecture | +| fe-developer | Frontend development | +| fe-qa | Frontend QA | --- ## Command Categories -### Workflow Commands +### Root Commands -High-level orchestration for multi-phase development: - -| Command | Purpose | -|---------|---------| -| `/workflow:plan` | Complete planning with session | -| `/workflow:lite-plan` | Lightweight interactive planning | -| `/workflow:lite-fix` | Bug diagnosis and fix | -| `/workflow:execute` | Execute tasks in session | -| `/workflow:resume` | Resume paused session | -| `/workflow:review` | Post-implementation review | -| `/workflow:status` | View workflow status | - -### Session Commands - -Session lifecycle management: - -| Command | Purpose | -|---------|---------| -| `/workflow:session:start` | Start new session | -| `/workflow:session:list` | List all sessions | -| `/workflow:session:resume` | Resume paused session | -| `/workflow:session:complete` | Archive completed session | -| `/workflow:session:sync` | Sync to project specs | - -### CLI Commands - -Direct AI tool access: - -| Command | Purpose | -|---------|---------| -| `/cli:cli-init` | Initialize CLI tool configs | -| `ccw cli -p "..." --tool gemini` | Execute with Gemini | -| `ccw cli -p "..." --tool codex --mode review` | Code review | - -### Memory Commands - -Context and documentation management: - -| Command | Purpose | -|---------|---------| -| `/memory:capture` | Capture session insights | -| `/memory:query` | Query captured memories | +| Command | Description | +|---------|-------------| +| `/ccw` | Main workflow orchestrator | +| `/ccw-coordinator` | Smart chain orchestrator | +| `/flow-create` | Flow template generator | ### Issue Commands -Post-development issue management: +| Command | Description | +|---------|-------------| +| `/issue/new` | Create new issue | +| `/issue/plan` | Batch plan issue resolution | +| `/issue/queue` | Form execution queue | +| `/issue/execute` | Execute queue | +| `/issue/discover` | Discover potential issues | +| `/issue/discover-by-prompt` | Discover from prompt | -| Command | Purpose | -|---------|---------| -| `/issue:discover` | Auto-discover issues | -| `/issue:discover-by-prompt` | Prompt-based discovery | -| `/issue:new` | Manual issue creation | -| `/issue:plan --all-pending` | Batch plan all pending | -| `/issue:queue` | Generate optimized queue | -| `/issue:execute` | Parallel execution | +### Workflow Commands + +| Command | Description | +|---------|-------------| +| `/workflow/init` | Initialize project state | +| `/workflow/init-specs` | Create spec files | +| `/workflow/init-guidelines` | Fill spec files | +| `/workflow/clean` | Code cleanup | +| `/workflow/analyze-with-file` | Collaborative analysis | +| `/workflow/brainstorm-with-file` | Brainstorming | +| `/workflow/collaborative-plan-with-file` | Collaborative planning | +| `/workflow/debug-with-file` | Debugging workflow | +| `/workflow/refactor-cycle` | Refactoring workflow | +| `/workflow/integration-test-cycle` | Integration testing | +| `/workflow/roadmap-with-file` | Roadmap planning | +| `/workflow/unified-execute-with-file` | Unified execution | + +### UI Design Commands + +| Command | Description | +|---------|-------------| +| `/workflow/ui-design/style-extract` | Extract styles | +| `/workflow/ui-design/layout-extract` | Extract layouts | +| `/workflow/ui-design/animation-extract` | Extract animations | +| `/workflow/ui-design/generate` | Generate UI prototypes | +| `/workflow/ui-design/import-from-code` | Import design from code | +| `/workflow/ui-design/codify-style` | Codify styles | +| `/workflow/ui-design/design-sync` | Sync design references | --- @@ -643,183 +230,134 @@ Post-development issue management: ### Workflow Skills -| Skill | Purpose | +| Skill | Trigger | |-------|---------| -| `workflow-plan` | Standard planning workflow | -| `workflow-lite-plan` | Lightweight planning | -| `workflow-multi-cli-plan` | Multi-CLI collaborative planning | -| `workflow-tdd` | Test-driven development | -| `workflow-test-fix` | Test fix generation | -| `workflow-execute` | Task execution engine | - -### Team Skills - -| Skill | Purpose | -|-------|---------| -| `team-lifecycle-v5` | Full lifecycle (spec -> impl -> test) | -| `team-coordinate-v2` | Team orchestration from scratch | -| `team-executor-v2` | Execute pre-planned team session | -| `team-brainstorm` | Multi-role brainstorming | -| `team-frontend` | Frontend-focused team | -| `team-testing` | Testing-focused team | -| `team-review` | Review-focused team | +| workflow-lite-plan | `workflow:lite-plan`, `workflow:lite-execute` | +| workflow-multi-cli-plan | `workflow:multi-cli-plan` | +| workflow-plan | `workflow:plan`, `workflow:plan-verify`, `workflow:replan` | +| workflow-execute | `workflow:execute` | +| workflow-tdd | `workflow:tdd-plan`, `workflow:tdd-verify` | +| workflow-test-fix | `workflow:test-fix-gen`, `workflow:test-cycle-execute` | ### Specialized Skills -| Skill | Purpose | +| Skill | Trigger | |-------|---------| -| `brainstorm` | Single-agent brainstorming | -| `review-cycle` | Code review cycle | -| `review-code` | Code review | -| `spec-generator` | Specification generation | -| `skill-generator` | Create new skills | -| `command-generator` | Generate slash commands | +| brainstorm | `brainstorm` | +| review-code | `review code` | +| review-cycle | `workflow:review-cycle` | +| spec-generator | `workflow:spec`, `generate spec` | +| skill-generator | `create skill` | +| skill-tuning | `skill tuning` | + +### Memory Skills + +| Skill | Trigger | +|-------|---------| +| memory-capture | `memory capture` | +| memory-manage | `memory manage` | --- ## Workflow Selection Guide -### Quick Selection Table - -| Scenario | Recommended Workflow | Level | -|----------|---------------------|-------| -| Quick fixes, config adjustments | `lite-lite-lite` | 1 | -| Clear single-module features | `lite-plan -> lite-execute` | 2 | -| Bug diagnosis and fix | `lite-fix` | 2 | -| Production emergencies | `lite-fix --hotfix` | 2 | -| Technology selection, solution comparison | `multi-cli-plan -> lite-execute` | 2 | -| Multi-module changes, refactoring | `plan -> verify -> execute` | 3 | -| Test-driven development | `tdd-plan -> execute -> tdd-verify` | 3 | -| Test failure fixes | `test-fix-gen -> test-cycle-execute` | 3 | -| New features, architecture design | `brainstorm:auto-parallel -> plan` | 4 | -| Complex multi-role projects | `team-lifecycle-v5` | Team | -| Post-development issue fixes | Issue Workflow | - | +``` + Task Complexity + Low Medium High + │ │ │ +────────────────────┌────────────┌─────────────┌──────────── + │ │ │ +Quick Fix │ │ │ +Config Change ───┌──> │ │ +Single Module │ │ │ + │ lite-plan │ │ + │ │ │ +────────────────────┌────────────┌─────────────┌──────────── + │ │ │ +Multi-Module │ │ │ +Feature Dev ─┌────────────┌──> │ + │ │ plan │ + │ │ │ +────────────────────┌────────────┌─────────────┌──────────── + │ │ │ +Architecture │ │ │ +New System ─┌────────────┌─────────────┌──> + │ │ │ brainstorm + │ │ │ + plan + │ │ │ + execute +``` ### Decision Flowchart ``` Start - | - +-- Post-development maintenance? - | +-- Yes -> Issue Workflow - | +-- No --> - | - +-- Need full team orchestration? - | +-- Yes -> team-lifecycle-v5 / team-coordinate-v2 - | +-- No --> - | - +-- Requirements clear? - | +-- Uncertain -> Level 4 (brainstorm:auto-parallel) - | +-- Clear --> - | - +-- Need persistent Session? - | +-- Yes -> Level 3 (plan / tdd-plan / test-fix-gen) - | +-- No --> - | - +-- Need multi-perspective / comparison? - | +-- Yes -> Level 2 (multi-cli-plan) - | +-- No --> - | - +-- Bug fix? - | +-- Yes -> Level 2 (lite-fix) - | +-- No --> - | - +-- Need planning? - +-- Yes -> Level 2 (lite-plan) - +-- No -> Level 1 (lite-lite-lite) + │ + ├─ Is it a quick fix or config change? + │ └─> Yes: workflow:lite-plan + │ + ├─ Is it a single module feature? + │ └─> Yes: workflow:lite-plan + │ + ├─ Does it need multi-CLI analysis? + │ └─> Yes: workflow:multi-cli-plan + │ + ├─ Is it multi-module with session? + │ └─> Yes: workflow:plan + │ + ├─ Is it TDD development? + │ └─> Yes: workflow:tdd-plan + │ + ├─ Is it test generation? + │ └─> Yes: workflow:test-fix-gen + │ + └─ Is it architecture/new system? + └─> Yes: brainstorm + workflow:plan ``` -### Complexity Indicators - -System auto-evaluates complexity: - -| Weight | Keywords | -|--------|----------| -| +2 | refactor, migrate, architect, system | -| +2 | multiple, across, all, entire | -| +1 | integrate, api, database | -| +1 | security, performance, scale | - -- **High complexity** (>=4): Auto-select Level 3/Team -- **Medium complexity** (2-3): Auto-select Level 2 -- **Low complexity** (<2): Auto-select Level 1 - --- ## Issue Workflow -**Main Workflow Supplement - Post-development continuous maintenance** - -### Two-Phase Lifecycle +### Issue Lifecycle ``` -+-------------------------------------------------------------+ -| Phase 1: Accumulation | -| | -| Triggers: | -| * Post-task review | -| * Code review findings | -| * Test failures | -| | -| +------------+ +------------+ +------------+ | -| | discover | | discover- | | new | | -| | Auto | | by-prompt | | Manual | | -| +------------+ +------------+ +------------+ | -| | -| Continuously accumulate issues to pending queue | -+-------------------------------------------------------------+ - | - | After sufficient accumulation - v -+-------------------------------------------------------------+ -| Phase 2: Batch Resolution | -| | -| +------------+ +------------+ +------------+ | -| | plan | -> | queue | -> | execute | | -| | --all- | | Optimize | | Parallel | | -| | pending | | order | | execution | | -| +------------+ +------------+ +------------+ | -| | -| Supports worktree isolation, maintains main branch stable | -+-------------------------------------------------------------+ +/issue/new Create issue with solution + ↓ +/issue/plan Batch plan resolution + ↓ +/issue/queue Form execution queue (DAG) + ↓ +/issue/execute Execute with parallel orchestration ``` -### Command List +### Issue Commands -**Accumulation Phase**: -```bash -/issue:discover # Multi-perspective auto-discovery -/issue:discover-by-prompt # Prompt-based discovery -/issue:new # Manual creation -``` - -**Batch Resolution**: -```bash -/issue:plan --all-pending # Batch plan all pending -/issue:queue # Generate optimized execution queue -/issue:execute # Parallel execution -``` +| Command | Purpose | +|---------|---------| +| `/issue/new` | Create structured issue from URL or description | +| `/issue/discover` | Discover issues from multiple perspectives | +| `/issue/plan` | Batch plan using issue-plan-agent | +| `/issue/queue` | Form queue using issue-queue-agent | +| `/issue/execute` | DAG-based parallel execution | --- -## Summary +## Quick Reference -### Level Overview +### Most Common Skills -| Level | Name | Workflows | Artifacts | Execution | -|-------|------|-----------|-----------|-----------| -| **1** | Rapid | `lite-lite-lite` | None | Direct | -| **2** | Lightweight | `lite-plan`, `lite-fix`, `multi-cli-plan` | Memory/Light | `lite-execute` | -| **3** | Standard | `plan`, `tdd-plan`, `test-fix-gen` | Session | `execute` | -| **4** | Brainstorm | `brainstorm:auto-parallel` | Multi-role + Session | `plan -> execute` | -| **Team** | Orchestration | `team-lifecycle-v5`, `team-*` | Full team artifacts | Coordinator | -| **Issue** | Maintenance | `discover -> plan -> queue -> execute` | Issue records | Worktree (optional) | +| Skill | When to Use | +|-------|-------------| +| `workflow:lite-plan` | Quick fixes, single features | +| `workflow:plan` | Multi-module development | +| `brainstorm` | Architecture, new features | +| `workflow:execute` | Execute planned work | -### Core Principles +### Most Common Commands -1. **Main Workflow** solves parallelism through **dependency analysis + Agent parallel execution** -2. **Issue Workflow** serves as **supplementary mechanism**, supports worktree isolation -3. Select appropriate workflow level based on task complexity, **avoid over-engineering** -4. **Levels 1-4** are manual command selection; **Team Architecture** provides intelligent orchestration -5. Use **Session Lifecycle** commands for full workflow state management -6. Leverage **Inner Loop** roles for batch processing of same-type tasks +| Command | When to Use | +|---------|-------------| +| `/ccw` | Auto workflow selection | +| `/workflow/session:start` | Start new session | +| `/workflow/session:resume` | Continue paused work | +| `/issue/new` | Create new issue | diff --git a/WORKFLOW_GUIDE_CN.md b/WORKFLOW_GUIDE_CN.md index c8ac13fc..8aa42a32 100644 --- a/WORKFLOW_GUIDE_CN.md +++ b/WORKFLOW_GUIDE_CN.md @@ -2,36 +2,12 @@ ## 抂述 -CCW 提䟛了基于 **团队架构 v2** 和 **4级工䜜流系统** 的完敎工䜜流䜓系芆盖从快速原型到完敎团队猖排的蜯件匀发党生呜呚期。 - -``` -+---------------------------------------------------------------+ -| 4级工䜜流系统 | -| | -| +----------+ +----------+ +----------+ +--------------+ | -| | Level 1 | ->| Level 2 | ->| Level 3 | ->| Level 4 | | -| | 急速执行 | | 蜻量规划 | | 标准规划 | | 倎脑风暎 | | -| | | | | | | | | | -| | lite- | |lite-plan | | plan | | brainstorm | | -| | lite-lite| |lite-fix | | tdd-plan | | :auto- | | -| | | |multi-cli-| |test-fix- | | parallel | | -| | | | plan | | gen | | | | | -| +----------+ +----------+ +----------+ +-----|--------+ | -| | | -| +------------v------+ | -| | 团队架构 v2 | | -| | team-coordinate | | -| | team-executor | | -| +-------------------+ | -| | -| 手劚皋床: 高 -------------------------------------> 䜎: 自劚 | -+---------------------------------------------------------------+ -``` +CCW 提䟛了基于 **Team 架构 v2** 和 **Skill 工䜜流系统** 的完敎工䜜流䜓系芆盖从快速原型到完敎团队猖排的蜯件匀发党生呜呚期。 ## v7.0 新增功胜 **䞻芁新特性**: -- **团队架构 v2**: `team-coordinate-v2` 和 `team-executor-v2` 统䞀 team-worker 代理 +- **Team 架构 v2**: `team-coordinate-v2` 和 `team-executor-v2` 统䞀 team-worker 代理 - **team-lifecycle-v5**: 完敎生呜呚期统䞀团队技胜 (规栌 -> 实现 -> 测试 -> 审查) - **队列调床噚**: 具有䟝赖解析的后台任务执行 - **工䜜流䌚话呜什**: `start`、`resume`、`complete`、`sync` 完敎生呜呚期管理 @@ -40,785 +16,349 @@ CCW 提䟛了基于 **团队架构 v2** 和 **4级工䜜流系统** 的完敎工 --- -## 目圕 +## Skills 侎 Commands -1. [4级工䜜流系统](#4级工䜜流系统) -2. [䌚话生呜呚期](#䌚话生呜呚期) -3. [团队架构 v2](#团队架构-v2) -4. [呜什分类](#呜什分类) -5. [技胜分类](#技胜分类) -6. [工䜜流选择指南](#工䜜流选择指南) -7. [Issue 工䜜流](#issue-工䜜流) +CCW 䜿甚䞀种调甚方匏 + +| 类型 | 栌匏 | 瀺䟋 | +|------|------|------| +| **Skills** | 觊发短语无斜杠 | `workflow:lite-plan`, `brainstorm`, `workflow:plan` | +| **Commands** | 斜杠呜什 | `/ccw`, `/workflow/session:start`, `/issue/new` | --- -## 4级工䜜流系统 +## 工䜜流 Skills -### Level 1: 急速执行 (lite-lite-lite) +### 蜻量规划 -**最快 - 盎接执行最小匀销** +| Skill 觊发词 | 甹途 | 阶段 | +|--------------|------|------| +| `workflow:lite-plan` | 蜻量规划䞎探玢 | 5 阶段 | +| `workflow:lite-execute` | 执行 lite-plan 蟓出 | 执行 | +**5 阶段亀互匏工䜜流** ``` -+-----------------+ +------------------+ -| lite-lite-lite | --> | 盎接执行 | -+-----------------+ +------------------+ +阶段 1: 任务分析䞎探玢 (30-90秒) +阶段 2: 柄枅甚户参䞎 +阶段 3: 规划 (20-60秒) +阶段 4: 䞉绎确讀 +阶段 5: 执行䞎跟螪 ``` -**特点**: -| 属性 | 倌 | -|------|-----| -| **倍杂床** | 最䜎 | -| **规划** | 内存䞭即时 | -| **产物** | 无仅代码 | -| **适甹** | 快速修倍、配眮调敎 | +### 倚 CLI 规划 -**呜什**: -```bash -/workflow:lite-lite-lite "䞺 API 添加 CORS 倎" +| Skill 觊发词 | 甹途 | +|--------------|------| +| `workflow:multi-cli-plan` | 倚 CLI 协同分析 | + +**5 阶段工䜜流** +``` +阶段 1: 䞊䞋文收集ACE 语义搜玢 +阶段 2: 倚 CLI 讚论迭代 +阶段 3: 展瀺选项 +阶段 4: 甚户决策 +阶段 5: 规划生成 ``` ---- +### 标准规划 -### Level 2: 蜻量规划 +| Skill 觊发词 | 甹途 | 阶段 | +|--------------|------|------| +| `workflow:plan` | 完敎规划䞎䌚话 | 5 阶段 | +| `workflow:plan-verify` | 规划验证 | 验证 | +| `workflow:replan` | 亀互匏重新规划 | 重规划 | -**快速 - 内存规划或单次分析快速迭代** +### TDD 工䜜流 -| 工䜜流 | 甹途 | 产物 | 执行方匏 | -|--------|------|------|----------| -| `lite-plan` | 明确需求的功胜匀发 | memory://plan | -> `lite-execute` | -| `lite-fix` | Bug 诊断修倍 | `.workflow/.lite-fix/` | -> `lite-execute` | -| `multi-cli-plan` | 需芁倚视角的任务 | `.workflow/.multi-cli-plan/` | -> `lite-execute` | +| Skill 觊发词 | 甹途 | +|--------------|------| +| `workflow:tdd-plan` | TDD 规划 | +| `workflow:tdd-verify` | TDD 验证 | -#### lite-plan -> lite-execute - -**5阶段亀互匏工䜜流**: +**6 阶段 TDD 规划 + Red-Green-Refactor** ``` -阶段1: 任务分析䞎探玢 (30-90秒) - └─ 自劚检测代码库䞊䞋文需求启劚 @cli-explore-agent - -阶段2: 柄枅 (甚户䟝赖) - └─ 基于探玢发现的亀互匏问答 - -阶段3: 规划 (20-60秒) - ├─ 䜎倍杂床 -> 盎接规划 - └─ äž­/高倍杂床 -> @cli-planning-agent - -阶段4: 䞉绎确讀 - └─ 任务批准 + 执行方匏 + 代码审查工具 - -阶段5: 执行䞎跟螪 (5-120分钟) - └─ 所选方法的实时进床 +阶段 1: 测试讟计 +阶段 2: Red猖写倱莥测试 +阶段 3: Green最小实现 +阶段 4: Refactor重构 +阶段 5: 验证 +阶段 6: 䞋䞀埪环 ``` -**关键特性**: -- 智胜代码探玢自劚检测或甚 `-e` 区制 -- 内存规划无文件产物 -- 䞉绎倚选确讀 -- 灵掻执行Agent 或 CLI -- 可选后续审查 +### 测试修倍工䜜流 -```bash -/workflow:lite-plan "添加 JWT 讀证" # 基本甚法 -/workflow:lite-plan -e "重构日志暡块" # 区制探玢 -/workflow:lite-execute # 执行计划 -``` +| Skill 觊发词 | 甹途 | +|--------------|------| +| `workflow:test-fix-gen` | 测试生成䞎修倍 | +| `workflow:test-cycle-execute` | 执行测试埪环 | -#### lite-fix +**析进匏测试层级 (L0-L3)** -**智胜 Bug 诊断 + 修倍5阶段**: -``` -阶段1: Bug 分析 - ├─ 䞥重性预评䌰 (Low/Medium/High/Critical) - └─ 并行 cli-explore-agent 诊断 (1-4 䞪角床) - -阶段2: 柄枅 (可选) - └─ 聚合柄枅需求AskUserQuestion - -阶段3: 修倍规划 - ├─ Low/Medium -> Claude 盎接规划 - └─ High/Critical -> cli-lite-planning-agent - -阶段4: 确讀 - └─ 甚户确讀执行方匏 - -阶段5: 执行 - └─ /workflow:lite-execute --in-memory --mode bugfix -``` - -```bash -/workflow:lite-fix # 标准修倍 -/workflow:lite-fix --hotfix # 玧急热修倍跳过诊断 -``` - -**产物**: `.workflow/.lite-fix/{bug-slug}-{date}/` - -#### multi-cli-plan - -**倚 CLI 协䜜分析 + 共识收敛5阶段**: -``` -阶段1: 䞊䞋文收集 - └─ ACE 语义搜玢构建䞊䞋文包 - -阶段2: 倚 CLI 讚论迭代 - ├─ cli-discuss-agent: Gemini + Codex + Claude - ├─ 亀叉验证合成方案 - └─ 埪环盎到收敛或蟟到最倧蜮数 - -阶段3: 展瀺选项 - └─ 展瀺方案及权衡 - -阶段4: 甚户决策 - └─ 甚户选择方案 - -阶段5: 计划生成 - └─ -> lite-execute -``` - -```bash -/workflow:multi-cli-plan "比蟃 OAuth 侎 JWT 方案" -``` - ---- - -### Level 3: 标准规划 - -**完敎 - 持久化 Session + 验证 + 完敎执行** - -| 工䜜流 | 甹途 | 阶段数 | 产物䜍眮 | -|--------|------|--------|----------| -| `plan` | 倍杂功胜匀发 | 5 阶段 | `.workflow/active/{session}/` | -| `tdd-plan` | 测试驱劚匀发 | 6 阶段 | `.workflow/active/{session}/` | -| `test-fix-gen` | 测试修倍生成 | 5 阶段 | `.workflow/active/WFS-test-{session}/` | - -#### plan -> verify -> execute - -**5阶段完敎规划**: -``` -阶段1: 䌚话发现 - └─ /workflow:session:start --auto - -阶段2: 䞊䞋文收集 - └─ /workflow:tools:context-gather - └─ 返回 context-package.json + conflict_risk - -阶段3: 冲突解决条件觊发 - └─ IF conflict_risk >= medium -> /workflow:tools:conflict-resolution - -阶段4: 任务生成 - └─ /workflow:tools:task-generate-agent - └─ 返回 IMPL_PLAN.md + IMPL-*.json + TODO_LIST.md - -阶段5: 摘芁 + 后续步骀 -``` - -```bash -/workflow:plan "倚暡块重构" # 完敎规划 -/workflow:plan-verify # 验证计划掚荐 -/workflow:execute # 执行 -/workflow:review # (可选) 审查 -``` - -**产物**: `.workflow/active/{WFS-session}/` -- `workflow-session.json` -- `IMPL_PLAN.md` -- `TODO_LIST.md` -- `.task/IMPL-*.json` - -#### tdd-plan -> execute -> tdd-verify - -**6阶段测试驱劚匀发**: -``` -阶段1: 䌚话发现 (--type tdd) -阶段2: 䞊䞋文收集 -阶段3: 测试芆盖率分析 -阶段4: 冲突解决条件觊发 -阶段5: TDD 任务生成Red-Green-Refactor 埪环 -阶段6: TDD 结构验证 -``` - -```bash -/workflow:tdd-plan "甚户讀证" -/workflow:execute -/workflow:tdd-verify -``` - -**TDD 任务结构**: -- 每䞪 IMPL 任务包含完敎的 Red-Green-Refactor 内郚埪环 -- `meta.tdd_workflow: true` -- `flow_control.implementation_approach` 包含 3 æ­¥ (red/green/refactor) - -#### test-fix-gen -> test-cycle-execute - -**5阶段测试修倍生成**: - -**双暡匏支持**: -| 暡匏 | 蟓入 | 䞊䞋文来源 | -|------|------|-----------| -| Session 暡匏 | `WFS-xxx` | 源 session 摘芁 | -| Prompt 暡匏 | 文本/文件路埄 | 盎接代码库分析 | - -```bash -/workflow:test-fix-gen WFS-user-auth-v2 # Session 暡匏 -/workflow:test-fix-gen "测试讀证 API" # Prompt 暡匏 -/workflow:test-cycle-execute # 执行测试修倍埪环 -``` - ---- - -### Level 4: 倎脑风暎 (brainstorm:auto-parallel) - -**探玢性 - 倚角色倎脑风暎 + 完敎规划** - -**3阶段流皋**: -``` -阶段1: 亀互匏框架生成 - └─ /workflow:brainstorm:artifacts - ├─ 䞻题分析生成问题 - ├─ 角色选择甚户确讀 - ├─ 角色问题收集 - └─ 生成 guidance-specification.md - -阶段2: 并行角色分析 - └─ N x Task(conceptual-planning-agent) - ├─ 每䞪角色独立分析 - └─ 并行生成 {role}/analysis.md - -阶段3: 绌合敎合 - └─ /workflow:brainstorm:synthesis - └─ 敎合所有分析 -> synthesis-specification.md -``` - -```bash -/workflow:brainstorm:auto-parallel "实时协䜜系统" -/workflow:plan --session {sessionId} -/workflow:execute -``` - -**可甚角色**: -| 角色 | 描述 | -|------|------| -| `system-architect` | 系统架构垈 | -| `ui-designer` | UI 讟计垈 | -| `ux-expert` | UX 䞓家 | -| `product-manager` | 产品经理 | -| `data-architect` | 数据架构垈 | -| `test-strategist` | 测试策略垈 | - -**产物结构**: -``` -.workflow/active/WFS-{topic}/ -├── workflow-session.json -└── .brainstorming/ - ├── guidance-specification.md - ├── {role}/analysis.md - └── synthesis-specification.md -``` +| 层级 | 名称 | 焊点 | +|------|------|------| +| **L0** | 静态分析 | 猖译、富入、类型、AI 代码问题 | +| **L1** | 单元测试 | 凜数/类行䞺 | +| **L2** | 集成测试 | 组件亀互、API 契纊 | +| **L3** | E2E 测试 | 甚户旅皋、关键路埄 | --- ## 䌚话生呜呚期 -CCW v7.0 匕入完敎的䌚话生呜呚期呜什甚于管理工䜜流䌚话从创建到完成的党过皋。 - -### 䌚话呜什抂览 - -| 呜什 | 甹途 | 䜿甚时机 | -|------|------|----------| -| `/workflow:session:start` | 启劚新䌚话或发现现有 | 匀始任䜕工䜜流 | -| `/workflow:session:resume` | 恢倍暂停的䌚话 | 返回䞭断的工䜜 | -| `/workflow:session:complete` | 園档䌚话并提取经验 | 所有任务完成后 | -| `/workflow:session:sync` | 同步䌚话工䜜到规范 | 曎新项目文档 | - -### 启劚䌚话 +### 䌚话呜什 ```bash -# 发现暡匏 - 列出掻劚䌚话 -/workflow:session:start - -# 自劚暡匏 - 智胜创建或重甚 -/workflow:session:start --auto "实现 OAuth2" - -# 区制新暡匏 -/workflow:session:start --new "甚户讀证功胜" - -# 指定䌚话类型 -/workflow:session:start --type tdd --auto "测试驱劚的登圕" +/workflow:session:start # 启劚新工䜜流䌚话 +/workflow:session:resume # 恢倍暂停的䌚话 +/workflow:session:list # 列出所有䌚话 +/workflow:session:sync # 同步䌚话工䜜 +/workflow:session:complete # 完成䌚话 +/workflow:session:solidify # 将孊习结晶䞺氞久记忆 ``` -**䌚话类型**: -- `workflow`: 标准实现默讀 -- `review`: 代码审查䌚话 -- `tdd`: 测试驱劚匀发 -- `test`: 测试生成/修倍䌚话 -- `docs`: 文档䌚话 +### 䌚话类型 -### 完成䌚话 - -```bash -/workflow:session:complete # 亀互匏完成 -/workflow:session:complete --yes # 自劚完成并同步 -/workflow:session:complete --detailed # 垊指标 -``` - -**完成操䜜**: -- 将䌚话園档到 `.workflow/archives/` -- 生成垊指标的 `manifest.json` -- 提取经验教训 -- 自劚同步项目状态䜿甚 `--yes` +| 类型 | 前猀 | 描述 | +|------|------|------| +| **工䜜流** | `WFS-` | 通甚匀发䌚话 | +| **审查** | `WFS-review-` | 代码审查䌚话 | +| **TDD** | `WFS-tdd-` | TDD 工䜜流䌚话 | +| **测试** | `WFS-test-` | 测试生成䌚话 | ### 䌚话目圕结构 ``` -.workflow/ -├── active/ # 掻劚䌚话 -│ └── WFS-{session-name}/ -│ ├── workflow-session.json # 䌚话元数据 -│ ├── IMPL_PLAN.md # 实现计划 -│ ├── TODO_LIST.md # 任务枅单 -│ ├── .task/ # 任务 JSON 文件 -│ └── .process/ # 过皋工件 -├── archives/ # 已完成的䌚话 -└── project-tech.json # 项目技术泚册衚 +.workflow/active/{session-id}/ +├── workflow-session.json # 䌚话元数据 +├── IMPL_PLAN.md # 实现计划 +├── TODO_LIST.md # 任务枅单 +├── .task/ # 任务 JSON 文件 +└── .process/ # 过皋产物 ``` --- -## 团队架构 v2 +## Team 架构 v2 -**适甚于需芁倚角色䞓䞚知识和猖排的倍杂项目。** +### 栞心抂念 -### 抂述 +- **team-worker 智胜䜓**: 所有角色的统䞀工䜜者智胜䜓 +- **role-spec 文件**: 蜻量级 YAML frontmatter + 阶段 2-4 逻蟑 +- **内埪环框架**: 同前猀任务的批倄理 +- **节拍/节奏暡型**: 事件驱劚协调 -团队架构 v2 (`team-coordinate-v2`、`team-executor-v2`、`team-lifecycle-v5`) 䞺倍杂蜯件匀发工䜜流提䟛统䞀的 team-worker 代理架构。 +### 可甚团队 Skills -``` -+---------------------------------------------------------------+ -| Team Coordinate / Team Executor v2 | -| | -| +-------------+ +-------------------------------------+ | -| | Coordinator | ---> | 劚态角色规范生成 | | -| | / Executor | | (analyst, planner, executor, etc.) | | -| +-------------+ +-------------------------------------+ | -| | | | -| v v | -| +-------------+ +-------------------------------------+ | -| | 任务 | | team-worker 代理 | | -| | 分发 | | 阶段1: 任务发现 (内眮) | | -| +-------------+ | 阶段2-4: 角色特定 (规范文件) | | -| | 阶段5: 报告 (内眮) | | -| +-------------------------------------+ | -| | | -| v | -| +-------------------------------------+ | -| | 子代理 (Discuss, Explore, Docs) | | -| +-------------------------------------+ | -+---------------------------------------------------------------+ -``` - -### team-worker 代理 - -统䞀的工䜜代理: -- **阶段1内眮**: 任务发现 - 按前猀和状态过滀任务 -- **阶段2-4角色特定**: 从角色规范 markdown 文件加蜜领域逻蟑 -- **阶段5内眮**: 报告 + 快速掚进 - 倄理完成和后继生成 - -### 角色规范文件 - -垊有 YAML 前眮信息的蜻量级 markdown 文件: - -```yaml ---- -role: analyst -prefix: RESEARCH -inner_loop: false -subagents: [explore, discuss] -message_types: - success: research_ready - error: error ---- -``` +| Skill | 甹途 | +|-------|------| +| `team-coordinate-v2` | 劚态角色生成䞎协调 | +| `team-executor-v2` | 现有䌚话的纯执行 | +| `team-lifecycle-v5` | 完敎生呜呚期规栌 -> 实现 -> 测试 | +| `team-brainstorm` | 倎脑风暎团队 | +| `team-frontend` | 前端匀发团队 | +| `team-testing` | 测试团队 | +| `team-review` | 代码审查团队 | ### 可甚角色 -| 角色 | 前猀 | 职莣 | 内埪环 | -|------|------|------|--------| -| analyst | RESEARCH | 代码库探玢、分析 | 吊 | -| writer | DRAFT | 文档生成 | 是 | -| planner | PLAN | 任务分解、䟝赖规划 | 是 | -| executor | IMPL | 实现和猖码 | 是 | -| tester | TEST | 测试和莚量保证 | 是 | -| reviewer | REVIEW | 代码审查和莚量闚 | 吊 | -| architect | ARCH | 架构决策 | 吊 | -| fe-developer | FE-IMPL | 前端实现 | 吊 | -| fe-qa | FE-TEST | 前端测试 | 吊 | - -### 内埪环框架 - -圓 `inner_loop: true` 时单䞪代理顺序倄理所有盞同前猀任务: - -``` -context_accumulator = [] - -阶段1: 查扟第䞀䞪 IMPL-* 任务 - 阶段2-4: 执行角色规范 - 阶段5-L: 标记完成记圕环积 - 曎倚 IMPL-* 任务 -> 阶段1埪环 - 没有了 -> 阶段5-F最终报告 -``` - -### 节拍/韵埋猖排暡型 - -**事件驱劚的协调暡型**: - -``` -节拍埪环单䞪节拍 -====================================================================== - 事件 协调噚 工䜜者 ----------------------------------------------------------------------- - 回调/恢倍 --> +- 倄理回调 -+ - | 标记完成 | - | 检查管道 | - +- 倄理生成䞋䞀䞪 -+ - | 查扟就绪任务 | - | 生成工䜜者 ---+--> [team-worker A] - | (可并行) ----+--> [team-worker B] - +- 停止 (空闲) ---+ | - | - 回调 <---------------------------------------+ - (䞋䞀䞪节拍) -====================================================================== -``` - -### 呜什 - -#### Team Coordinate - -从倎生成角色规范并猖排团队: - -```bash -/team-coordinate "讟计和实现实时协䜜系统" -``` - -**流皋**: -1. 分析需求并检测胜力 -2. 劚态生成角色规范 -3. 创建垊䟝赖铟的任务 -4. 生成 team-worker 代理 -5. 通过回调监控进床 -6. 完成并生成绌合报告 - -#### Team Executor - -执行预规划的团队䌚话: - -```bash -/team-executor # 初始执行 -/team-executor resume # 恢倍暂停的䌚话 -/team-executor status # 检查状态 -``` - -#### Team Lifecycle v5 - -完敎生呜呚期统䞀团队技胜规栌 -> 实现 -> 测试 -> 审查: - -```bash -# 觊发词 "team lifecycle" -/team-lifecycle "构建垊 OAuth2 的甚户讀证系统" -``` - -**管道定义**: - -| 管道 | 任务数 | 流皋 | -|------|--------|------| -| 仅规栌 | 6 | RESEARCH -> DRAFT-001..004 -> QUALITY | -| 仅实现 | 4 | PLAN -> IMPL -> TEST + REVIEW | -| 完敎生呜呚期 | 10 | [规栌] -> PLAN -> IMPL -> TEST + REVIEW | -| 前端 | 3+ | PLAN -> DEV-FE -> QA-FE (GC 埪环) | - -### 子代理 - -| 子代理 | 甹途 | -|--------|------| -| discuss | 倚视角批刀劚态视角 | -| explore | 代码库探玢垊猓存 | -| doc-generation | 从暡板生成文档 | - -### 消息总线协议 - -```javascript -mcp__ccw-tools__team_msg({ - operation: "log", - team: "", // 䌚话 ID䞍是团队名称 - from: "", - to: "coordinator", - type: "", - summary: "[] <消息>", - ref: "" -}) -``` - -### 䌚话结构 - -``` -.workflow/.team// -├── team-session.json # 䌚话元数据 -├── task-analysis.json # 任务䟝赖 -├── role-specs/ # 生成的角色规范文件 -├── artifacts/ # 任务蟓出 -├── discussions/ # 倚视角批刀 -├── explorations/ # 猓存的探玢结果 -└── wisdom/ # 环积的经验 - ├── learnings.md - ├── decisions.md - ├── conventions.md - └── issues.md -``` +| 角色 | 职莣 | +|------|------| +| analyst | 代码分析、需求 | +| writer | 文档、内容 | +| planner | 规划、架构 | +| executor | 实现、猖码 | +| tester | 测试、QA | +| reviewer | 代码审查、反銈 | +| architect | 系统讟计、架构 | +| fe-developer | 前端匀发 | +| fe-qa | 前端 QA | --- ## 呜什分类 -### 工䜜流呜什 +### 根呜什 -倚阶段匀发的高级猖排: - -| 呜什 | 甹途 | +| 呜什 | 描述 | |------|------| -| `/workflow:plan` | 垊䌚话的完敎规划 | -| `/workflow:lite-plan` | 蜻量亀互匏规划 | -| `/workflow:lite-fix` | Bug 诊断和修倍 | -| `/workflow:execute` | 执行䌚话䞭的任务 | -| `/workflow:resume` | 恢倍暂停的䌚话 | -| `/workflow:review` | 实现后审查 | -| `/workflow:status` | 查看工䜜流状态 | - -### 䌚话呜什 - -䌚话生呜呚期管理: - -| 呜什 | 甹途 | -|------|------| -| `/workflow:session:start` | 启劚新䌚话 | -| `/workflow:session:list` | 列出所有䌚话 | -| `/workflow:session:resume` | 恢倍暂停的䌚话 | -| `/workflow:session:complete` | 園档已完成的䌚话 | -| `/workflow:session:sync` | 同步到项目规范 | - -### CLI 呜什 - -盎接访问 AI 工具: - -| 呜什 | 甹途 | -|------|------| -| `/cli:cli-init` | 初始化 CLI 工具配眮 | -| `ccw cli -p "..." --tool gemini` | 䜿甚 Gemini 执行 | -| `ccw cli -p "..." --tool codex --mode review` | 代码审查 | - -### 内存呜什 - -䞊䞋文和文档管理: - -| 呜什 | 甹途 | -|------|------| -| `/memory:capture` | 捕获䌚话掞察 | -| `/memory:query` | 查询捕获的内存 | +| `/ccw` | 䞻工䜜流猖排噚 | +| `/ccw-coordinator` | 智胜铟猖排噚 | +| `/flow-create` | 流皋暡板生成噚 | ### Issue 呜什 -匀发后问题管理: - -| 呜什 | 甹途 | +| 呜什 | 描述 | |------|------| -| `/issue:discover` | 自劚发现问题 | -| `/issue:discover-by-prompt` | 基于提瀺发现 | -| `/issue:new` | 手劚创建问题 | -| `/issue:plan --all-pending` | 批量规划所有埅倄理 | -| `/issue:queue` | 生成䌘化队列 | -| `/issue:execute` | 并行执行 | +| `/issue/new` | 创建新 issue | +| `/issue/plan` | 批量规划 issue 解决方案 | +| `/issue/queue` | 圢成执行队列 | +| `/issue/execute` | 执行队列 | +| `/issue/discover` | 发现朜圚问题 | +| `/issue/discover-by-prompt` | 从提瀺发现 | + +### 工䜜流呜什 + +| 呜什 | 描述 | +|------|------| +| `/workflow/init` | 初始化项目状态 | +| `/workflow/init-specs` | 创建规栌文件 | +| `/workflow/init-guidelines` | 填充规栌文件 | +| `/workflow/clean` | 代码枅理 | +| `/workflow/analyze-with-file` | 协同分析 | +| `/workflow/brainstorm-with-file` | 倎脑风暎 | +| `/workflow/collaborative-plan-with-file` | 协同规划 | +| `/workflow/debug-with-file` | 调试工䜜流 | +| `/workflow/refactor-cycle` | 重构工䜜流 | +| `/workflow/integration-test-cycle` | 集成测试 | +| `/workflow/roadmap-with-file` | 路线囟规划 | +| `/workflow/unified-execute-with-file` | 统䞀执行 | + +### UI 讟计呜什 + +| 呜什 | 描述 | +|------|------| +| `/workflow/ui-design/style-extract` | 提取样匏 | +| `/workflow/ui-design/layout-extract` | 提取垃局 | +| `/workflow/ui-design/animation-extract` | 提取劚画 | +| `/workflow/ui-design/generate` | 生成 UI 原型 | +| `/workflow/ui-design/import-from-code` | 从代码富入讟计 | +| `/workflow/ui-design/codify-style` | 猖码样匏 | +| `/workflow/ui-design/design-sync` | 同步讟计匕甚 | --- -## 技胜分类 +## Skill 分类 -### 工䜜流技胜 +### 工䜜流 Skills -| 技胜 | 甹途 | -|------|------| -| `workflow-plan` | 标准规划工䜜流 | -| `workflow-lite-plan` | 蜻量规划 | -| `workflow-multi-cli-plan` | 倚 CLI 协䜜规划 | -| `workflow-tdd` | 测试驱劚匀发 | -| `workflow-test-fix` | 测试修倍生成 | -| `workflow-execute` | 任务执行匕擎 | +| Skill | 觊发词 | +|-------|--------| +| workflow-lite-plan | `workflow:lite-plan`, `workflow:lite-execute` | +| workflow-multi-cli-plan | `workflow:multi-cli-plan` | +| workflow-plan | `workflow:plan`, `workflow:plan-verify`, `workflow:replan` | +| workflow-execute | `workflow:execute` | +| workflow-tdd | `workflow:tdd-plan`, `workflow:tdd-verify` | +| workflow-test-fix | `workflow:test-fix-gen`, `workflow:test-cycle-execute` | -### 团队技胜 +### 䞓项 Skills -| 技胜 | 甹途 | -|------|------| -| `team-lifecycle-v5` | 完敎生呜呚期规栌 -> 实现 -> 测试 | -| `team-coordinate-v2` | 从倎猖排团队 | -| `team-executor-v2` | 执行预规划的团队䌚话 | -| `team-brainstorm` | 倚角色倎脑风暎 | -| `team-frontend` | 前端聚焊团队 | -| `team-testing` | 测试聚焊团队 | -| `team-review` | 审查聚焊团队 | +| Skill | 觊发词 | +|-------|--------| +| brainstorm | `brainstorm` | +| review-code | `review code` | +| review-cycle | `workflow:review-cycle` | +| spec-generator | `workflow:spec`, `generate spec` | +| skill-generator | `create skill` | +| skill-tuning | `skill tuning` | -### 䞓甚技胜 +### 记忆 Skills -| 技胜 | 甹途 | -|------|------| -| `brainstorm` | 单代理倎脑风暎 | -| `review-cycle` | 代码审查埪环 | -| `review-code` | 代码审查 | -| `spec-generator` | 规栌生成 | -| `skill-generator` | 创建新技胜 | -| `command-generator` | 生成斜杠呜什 | +| Skill | 觊发词 | +|-------|--------| +| memory-capture | `memory capture` | +| memory-manage | `memory manage` | --- ## 工䜜流选择指南 -### 快速选择衚 - -| 场景 | 掚荐工䜜流 | Level | -|------|-----------|-------| -| 快速修倍、配眮调敎 | `lite-lite-lite` | 1 | -| 明确需求的单暡块功胜 | `lite-plan -> lite-execute` | 2 | -| Bug 诊断修倍 | `lite-fix` | 2 | -| 玧急生产问题 | `lite-fix --hotfix` | 2 | -| 技术选型、方案比蟃 | `multi-cli-plan -> lite-execute` | 2 | -| 倚暡块改劚、重构 | `plan -> verify -> execute` | 3 | -| 测试驱劚匀发 | `tdd-plan -> execute -> tdd-verify` | 3 | -| 测试倱莥修倍 | `test-fix-gen -> test-cycle-execute` | 3 | -| 党新功胜、架构讟计 | `brainstorm:auto-parallel -> plan` | 4 | -| 倍杂倚角色项目 | `team-lifecycle-v5` | Team | -| 匀发后问题修倍 | Issue Workflow | - | +``` + 任务倍杂床 + 䜎 äž­ 高 + │ │ │ +────────────────────┌────────────┌─────────────┌──────────── + │ │ │ +快速修倍 │ │ │ +配眮变曎 ────┌──> │ │ +单暡块功胜 │ │ │ + │ lite-plan │ │ + │ │ │ +────────────────────┌────────────┌─────────────┌──────────── + │ │ │ +倚暡块功胜 │ │ │ +功胜匀发 ────┌────────────┌──> │ + │ │ plan │ + │ │ │ +────────────────────┌────────────┌─────────────┌──────────── + │ │ │ +架构讟计 │ │ │ +新系统 ────┌────────────┌─────────────┌──> + │ │ │ brainstorm + │ │ │ + plan + │ │ │ + execute +``` ### 决策流皋囟 ``` 匀始 - | - +-- 是匀发后的绎技问题 - | +-- 是 -> Issue Workflow - | +-- 吊 --> - | - +-- 需芁完敎团队猖排 - | +-- 是 -> team-lifecycle-v5 / team-coordinate-v2 - | +-- 吊 --> - | - +-- 需求是吊明确 - | +-- 䞍确定 -> Level 4 (brainstorm:auto-parallel) - | +-- 明确 --> - | - +-- 需芁持久化 Session - | +-- 是 -> Level 3 (plan / tdd-plan / test-fix-gen) - | +-- 吊 --> - | - +-- 需芁倚视角/方案比蟃 - | +-- 是 -> Level 2 (multi-cli-plan) - | +-- 吊 --> - | - +-- 是 Bug 修倍 - | +-- 是 -> Level 2 (lite-fix) - | +-- 吊 --> - | - +-- 是吊需芁规划 - +-- 是 -> Level 2 (lite-plan) - +-- 吊 -> Level 1 (lite-lite-lite) + │ + ├─ 是快速修倍或配眮变曎 + │ └─> 是workflow:lite-plan + │ + ├─ 是单暡块功胜 + │ └─> 是workflow:lite-plan + │ + ├─ 需芁倚 CLI 分析 + │ └─> 是workflow:multi-cli-plan + │ + ├─ 是倚暡块䞔需芁䌚话 + │ └─> 是workflow:plan + │ + ├─ 是 TDD 匀发 + │ └─> 是workflow:tdd-plan + │ + ├─ 是测试生成 + │ └─> 是workflow:test-fix-gen + │ + └─ 是架构/新系统 + └─> 是brainstorm + workflow:plan ``` -### 倍杂床指标 - -系统自劚评䌰倍杂床: - -| 权重 | 关键词 | -|------|--------| -| +2 | refactor, 重构, migrate, 迁移, architect, 架构, system, 系统 | -| +2 | multiple, 倚䞪, across, è·š, all, 所有, entire, 敎䞪 | -| +1 | integrate, 集成, api, database, 数据库 | -| +1 | security, 安党, performance, 性胜, scale, 扩展 | - -- **高倍杂床** (>=4): 自劚选择 Level 3/Team -- **䞭倍杂床** (2-3): 自劚选择 Level 2 -- **䜎倍杂床** (<2): 自劚选择 Level 1 - --- ## Issue 工䜜流 -**䞻干工䜜流的补充 - 匀发后的持续绎技** - -### 䞀阶段生呜呚期 +### Issue 生呜呚期 ``` -+-------------------------------------------------------------+ -| 阶段1: 积环 | -| | -| 觊发源: | -| * 任务完成后的 review | -| * 代码审查发现 | -| * 测试倱莥 | -| | -| +------------+ +------------+ +------------+ | -| | discover | | discover- | | new | | -| | 自劚 | | by-prompt | | 手劚创建 | | -| +------------+ +------------+ +------------+ | -| | -| 持续积环 Issue 到埅倄理队列 | -+-------------------------------------------------------------+ - | - | 积环足借后 - v -+-------------------------------------------------------------+ -| 阶段2: 批量解决 | -| | -| +------------+ +------------+ +------------+ | -| | plan | -> | queue | -> | execute | | -| | --all- | | 䌘化顺序 | | 并行执行 | | -| | pending | | 冲突分析 | | | | -| +------------+ +------------+ +------------+ | -| | -| 支持 Worktree 隔犻保持䞻分支皳定 | -+-------------------------------------------------------------+ +/issue/new 创建垊解决方案的 issue + ↓ +/issue/plan 批量规划解决方案 + ↓ +/issue/queue 圢成执行队列DAG + ↓ +/issue/execute 并行猖排执行 ``` -### 呜什枅单 +### Issue 呜什 -**积环阶段**: -```bash -/issue:discover # 倚视角自劚发现 -/issue:discover-by-prompt # 基于提瀺发现 -/issue:new # 手劚创建 -``` - -**批量解决**: -```bash -/issue:plan --all-pending # 批量规划所有埅倄理 -/issue:queue # 生成䌘化执行队列 -/issue:execute # 并行执行 -``` +| 呜什 | 甹途 | +|------|------| +| `/issue/new` | 从 URL 或描述创建结构化 issue | +| `/issue/discover` | 从倚视角发现问题 | +| `/issue/plan` | 䜿甚 issue-plan-agent 批量规划 | +| `/issue/queue` | 䜿甚 issue-queue-agent 圢成队列 | +| `/issue/execute` | 基于 DAG 的并行执行 | --- -## 总结 +## 快速参考 -### 分级总览 +### 最垞甚 Skills -| Level | 名称 | 工䜜流 | 产物 | 执行方匏 | -|-------|------|--------|------|----------| -| **1** | 急速 | `lite-lite-lite` | 无 | 盎接执行 | -| **2** | 蜻量 | `lite-plan`, `lite-fix`, `multi-cli-plan` | 内存/蜻量 | `lite-execute` | -| **3** | 标准 | `plan`, `tdd-plan`, `test-fix-gen` | Session | `execute` | -| **4** | 倎脑风暎 | `brainstorm:auto-parallel` | 倚角色 + Session | `plan -> execute` | -| **Team** | 猖排 | `team-lifecycle-v5`, `team-*` | 完敎团队产物 | Coordinator | -| **Issue** | 绎技 | `discover -> plan -> queue -> execute` | Issue 记圕 | Worktree (可选) | +| Skill | 䜕时䜿甚 | +|-------|----------| +| `workflow:lite-plan` | 快速修倍、单功胜 | +| `workflow:plan` | 倚暡块匀发 | +| `brainstorm` | 架构、新功胜 | +| `workflow:execute` | 执行已规划的工䜜 | -### 栞心原则 +### 最垞甚 Commands -1. **䞻干工䜜流**通过**䟝赖分析 + Agent 并行**解决并行问题 -2. **Issue 工䜜流**䜜䞺**补充机制**支持 worktree 隔犻 -3. 根据任务倍杂床选择合适的工䜜流层级**避免过床工皋化** -4. **Level 1-4** 是手劚呜什选择**团队架构**提䟛智胜猖排 -5. 䜿甚**䌚话生呜呚期**呜什进行完敎工䜜流状态管理 -6. 利甚**内埪环**角色进行同类型任务的批倄理 +| 呜什 | 䜕时䜿甚 | +|------|----------| +| `/ccw` | 自劚工䜜流选择 | +| `/workflow/session:start` | 启劚新䌚话 | +| `/workflow/session:resume` | 继续暂停的工䜜 | +| `/issue/new` | 创建新 issue |