thanh-kit 2.5.1 → 2.5.2

package/templates/router/commands-guide.md

@@ -1,122 +0,0 @@
-# Commands Guide
-
-> **Purpose**: Each command file contains step-by-step workflow for a specific task type.
-> **Location**: `.claude/commands/<command>.md` or `.claude/commands/<category>/<variant>.md`
-> **When to load**: After selecting an Agent, to know the exact steps to execute.
-
----
-
-## Main Commands
-
-| Trigger | Command | Load File |
-|:---|:---|:---|
-| code, write code, implement | `/code` | `.claude/commands/code.md` |
-| fix, fix bug, fix error | `/fix` | `.claude/commands/fix.md` |
-| test, run tests | `/test` | `.claude/commands/test.md` |
-| plan, make plan | `/plan` | `.claude/commands/plan.md` |
-| review, review code | `/review-changes` | `.claude/commands/review-changes.md` |
-| build, compile, package | `/build` | `.claude/commands/build.md` |
-| debug, investigate | `/debug` | `.claude/commands/debug.md` |
-| scout, search | `/scout` | `.claude/commands/scout.md` |
-| brainstorm | `/brainstorm` | `.claude/commands/brainstorm.md` |
-
----
-
-## Sub-commands (Specialized Variants)
-
-### Code Variants
-| Trigger | Load File |
-|:---|:---|
-| auto code, automatic | `.claude/commands/code/auto.md` |
-| code without tests | `.claude/commands/code/no-test.md` |
-| parallel coding | `.claude/commands/code/parallel.md` |
-
-### Fix Variants
-| Trigger | Load File |
-|:---|:---|
-| quick fix | `.claude/commands/fix/fast.md` |
-| hard fix, complex bug | `.claude/commands/fix/hard.md` |
-| fix UI issue | `.claude/commands/fix/ui.md` |
-| fix failing tests | `.claude/commands/fix/test.md` |
-| fix type errors, typescript | `.claude/commands/fix/types.md` |
-| fix CI/CD | `.claude/commands/fix/ci.md` |
-| fix from logs | `.claude/commands/fix/logs.md` |
-| parallel fixing | `.claude/commands/fix/parallel.md` |
-
-### Plan Variants
-| Trigger | Load File |
-|:---|:---|
-| quick plan | `.claude/commands/plan/fast.md` |
-| complex plan | `.claude/commands/plan/hard.md` |
-| parallel planning | `.claude/commands/plan/parallel.md` |
-| two-phase plan | `.claude/commands/plan/two.md` |
-| validate plan | `.claude/commands/plan/validate.md` |
-| archive plan | `.claude/commands/plan/archive.md` |
-| CRO plan | `.claude/commands/plan/cro.md` |
-
-### Git Commands
-| Trigger | Load File |
-|:---|:---|
-| commit | `.claude/commands/git/cm.md` |
-| cherry-pick | `.claude/commands/git/cp.md` |
-| merge | `.claude/commands/git/merge.md` |
-| PR, pull request | `.claude/commands/git/pr.md` |
-
-### Docs Commands
-| Trigger | Load File |
-|:---|:---|
-| init docs | `.claude/commands/docs/init.md` |
-| update docs | `.claude/commands/docs/update.md` |
-| summarize docs | `.claude/commands/docs/summarize.md` |
-
-### Design Commands
-| Trigger | Load File |
-|:---|:---|
-| quick design | `.claude/commands/design/fast.md` |
-| good design | `.claude/commands/design/good.md` |
-| design from screenshot | `.claude/commands/design/screenshot.md` |
-| design from video | `.claude/commands/design/video.md` |
-| 3D design | `.claude/commands/design/3d.md` |
-| describe design | `.claude/commands/design/describe.md` |
-
-### Review Commands
-| Trigger | Load File |
-|:---|:---|
-| review codebase | `.claude/commands/review/codebase.md` |
-| post-task review | `.claude/commands/review/post-task.md` |
-
-### Skill Management
-| Trigger | Load File |
-|:---|:---|
-| add skill | `.claude/commands/skill/add.md` |
-| create skill | `.claude/commands/skill/create.md` |
-| plan skill | `.claude/commands/skill/plan.md` |
-| optimize skill | `.claude/commands/skill/optimize.md` |
-
-### Other Commands
-| Trigger | Load File |
-|:---|:---|
-| bootstrap project | `.claude/commands/bootstrap.md` |
-| create feature | `.claude/commands/create-feature.md` |
-| db migrate | `.claude/commands/db-migrate.md` |
-| generate DTO | `.claude/commands/generate-dto.md` |
-| lint | `.claude/commands/lint.md` |
-| kanban | `.claude/commands/kanban.md` |
-| preview | `.claude/commands/preview.md` |
-| worktree | `.claude/commands/worktree.md` |
-| use MCP | `.claude/commands/use-mcp.md` |
-| cook (recipe) | `.claude/commands/cook.md` |
-| journal | `.claude/commands/journal.md` |
-| watzup (status) | `.claude/commands/watzup.md` |
-| coding level | `.claude/commands/coding-level.md` |
-| ask (Q&A) | `.claude/commands/ask.md` |
-| scout external | `.claude/commands/scout/ext.md` |
-| test UI | `.claude/commands/test/ui.md` |
-
----
-
-## Usage Rules
-
-1. **Prefer specific over generic**: Use `/fix/ui` instead of `/fix` when fixing UI bugs.
-2. **Command = Workflow**: The file contains exact steps to follow.
-3. **Combine with Agent**: Agent defines mindset, Command defines process.

package/templates/router/decision-flow.md

@@ -1,92 +0,0 @@
-# Decision Flow - How AI Routes Requests
-
-## Request Processing Pipeline
-
-```
-User Request
-    │
-    ▼
-┌─────────────────────────────────────────────────────────┐
-│ STEP 1: Analyze Request                                 │
-│ - Identify WORK TYPE (code/fix/test/plan/review...)     │
-│ - Identify DOMAIN (frontend/backend/database/devops...) │
-│ - Identify COMPLEXITY (simple/complex/multi-step)       │
-└─────────────────────────────────────────────────────────┘
-    │
-    ▼
-┌─────────────────────────────────────────────────────────┐
-│ STEP 2: Select Agent (PRIMARY ROLE)                     │
-│ - Simple task: Pick 1 most suitable Agent               │
-│ - Complex task: May need multiple Agents coordination   │
-│ → Load file `.claude/agents/<agent-name>.md` for mindset│
-│ → See: .claude/router/agents-guide.md                   │
-└─────────────────────────────────────────────────────────┘
-    │
-    ▼
-┌─────────────────────────────────────────────────────────┐
-│ STEP 3: Select Command (WORKFLOW)                       │
-│ - Identify main command (code/fix/test/plan...)         │
-│ - If more specific variant exists, use sub-command      │
-│ → Load file `.claude/commands/<command>.md` for steps   │
-│ → See: .claude/router/commands-guide.md                 │
-└─────────────────────────────────────────────────────────┘
-    │
-    ▼
-┌─────────────────────────────────────────────────────────┐
-│ STEP 4: Load Skill (EXPERT KNOWLEDGE) - IF NEEDED       │
-│ - Scan keywords in request                              │
-│ - Only load when deep expertise is required             │
-│ → Load file `.claude/skills/<skill>/SKILL.md`           │
-│ → See: .claude/router/skills-guide.md                   │
-└─────────────────────────────────────────────────────────┘
-    │
-    ▼
-┌─────────────────────────────────────────────────────────┐
-│ STEP 5: Apply Workflow (IF large/complex task)          │
-│ - New feature → primary-workflow.md                     │
-│ - Multi-agent → orchestration-protocol.md               │
-│ → Load file `.claude/workflows/<workflow>.md`           │
-│ → See: .claude/router/workflows-guide.md                │
-└─────────────────────────────────────────────────────────┘
-    │
-    ▼
-  EXECUTE
-```
-
----
-
-## Quick Decision Matrix
-
-| Situation | Agent | Command | Skill | Workflow |
-|:---|:---|:---|:---|:---|
-| Small bug, clear cause | Debugger | `/fix/fast` | ❌ | ❌ |
-| Hard bug, needs investigation | Debugger | `/fix/hard` | `debugging/` | ❌ |
-| New feature code | Developer | `/code` | By domain | `primary-workflow` |
-| Create unit tests | Tester | `/test` | `test-generation/` | ❌ |
-| Code review before merge | Reviewer | `/review-changes` | `code-review/` | ❌ |
-| Plan large task | Planner | `/plan` | `planning/` | ❌ |
-| Find file/function in repo | Scout | `/scout` | ❌ | ❌ |
-| Create PR with proper message | Git Manager | `/git/pr` | ❌ | ❌ |
-| Design UI from screenshot | UI/UX Designer | `/design/screenshot` | `frontend-design/` | ❌ |
-| Complex multi-step task | Planner → Developer → Tester | `/plan` → `/code` → `/test` | Multiple | `primary-workflow` |
-
----
-
-## Selection Principles
-
-1. **Agent**: Always pick 1 primary agent. Only coordinate multiple agents when truly needed.
-2. **Command**: Prefer specific sub-command over generic command (e.g., `/fix/ui` over `/fix`).
-3. **Skill**: Only load when deep expertise is required. Skip for simple tasks.
-4. **Workflow**: Only use for large, multi-step tasks requiring orchestration.
-
----
-
-## Example Routing
-
-| User Request | → Agent | → Command | → Skill | → Workflow |
-|:---|:---|:---|:---|:---|
-| "Fix UI bug" | `.claude/agents/debugger.md` | `.claude/commands/fix/ui.md` | ❌ | ❌ |
-| "Write tests for feature X" | `.claude/agents/tester.md` | `.claude/commands/test.md` | `.claude/skills/test-generation/` | ❌ |
-| "Create PR" | `.claude/agents/git-manager.md` | `.claude/commands/git/pr.md` | ❌ | ❌ |
-| "Optimize performance" | `.claude/agents/developer.md` | `.claude/commands/code.md` | `.claude/skills/arch-performance-optimization/` | ❌ |
-| "Plan microservice architecture" | `.claude/agents/planner.md` | `.claude/commands/plan.md` | `.claude/skills/arch-cross-service-integration/` | `.claude/workflows/primary-workflow` |

package/templates/router/skills-guide.md

@@ -1,127 +0,0 @@
-# Skills Guide
-
-> **Purpose**: Skills provide deep expert knowledge for specific domains/technologies.
-> **Location**: `.claude/skills/<skill-name>/SKILL.md`
-> **When to load**: Only when task requires specialized expertise. Skip for simple tasks.
-
----
-
-## Loading Rules
-
-1. **Scan keywords** in user request
-2. **Only load when needed**: Simple tasks don't need skills
-3. **Minimum set**: If multiple skills match, choose the smallest set that covers the request
-4. **Ask if ambiguous**: If unsure which skill, ask 1 clarifying question
-
----
-
-## Available Skills by Domain
-
-### Project Navigation
-| Trigger Keywords | Skill Folder |
-|:---|:---|
-| Project structure, index, tree, folder layout | `.claude/skills/project-index/` |
-
-### Frontend & UI
-| Trigger Keywords | Skill Folder |
-|:---|:---|
-| CSS, Tailwind, styling | `.claude/skills/ui-styling/` |
-| UI/UX advanced | `.claude/skills/ui-ux-pro-max/` |
-| Angular component | `.claude/skills/frontend-angular-component/` |
-| Angular form | `.claude/skills/frontend-angular-form/` |
-| Angular store | `.claude/skills/frontend-angular-store/` |
-| Angular API service | `.claude/skills/frontend-angular-api-service/` |
-| Frontend design | `.claude/skills/frontend-design/`, `.claude/skills/frontend-design-pro/` |
-| Frontend development | `.claude/skills/frontend-development/` |
-| Three.js, 3D | `.claude/skills/threejs/` |
-
-### Backend & Database
-| Trigger Keywords | Skill Folder |
-|:---|:---|
-| Backend, API | `.claude/skills/backend-development/` |
-| Database, SQL, query | `.claude/skills/databases/` |
-| EasyPlatform backend | `.claude/skills/easyplatform-backend/` |
-
-### Testing & Debugging
-| Trigger Keywords | Skill Folder |
-|:---|:---|
-| Test generation | `.claude/skills/test-generation/` |
-| Webapp testing | `.claude/skills/webapp-testing/` |
-| Bug diagnosis | `.claude/skills/bug-diagnosis/` |
-| Debugging | `.claude/skills/debugging/` |
-| Chrome DevTools | `.claude/skills/chrome-devtools/` |
-
-### Architecture & Performance
-| Trigger Keywords | Skill Folder |
-|:---|:---|
-| Cross-service, microservice | `.claude/skills/arch-cross-service-integration/` |
-| Performance optimization | `.claude/skills/arch-performance-optimization/` |
-| Security review | `.claude/skills/arch-security-review/` |
-
-### Documentation & Planning
-| Trigger Keywords | Skill Folder |
-|:---|:---|
-| Documentation | `.claude/skills/documentation/` |
-| Feature docs | `.claude/skills/feature-docs/` |
-| README improvement | `.claude/skills/readme-improvement/` |
-| Planning | `.claude/skills/planning/` |
-| Plan analysis | `.claude/skills/plan-analysis/` |
-| Plans kanban | `.claude/skills/plans-kanban/` |
-
-### AI & Media
-| Trigger Keywords | Skill Folder |
-|:---|:---|
-| AI multimodal, image, vision | `.claude/skills/ai-multimodal/` |
-| AI artist | `.claude/skills/ai-artist/` |
-| Media processing | `.claude/skills/media-processing/` |
-
-### DevOps & Tools
-| Trigger Keywords | Skill Folder |
-|:---|:---|
-| DevOps, CI/CD | `.claude/skills/devops/` |
-| MCP builder | `.claude/skills/mcp-builder/` |
-| MCP management | `.claude/skills/mcp-management/` |
-| Repomix | `.claude/skills/repomix/` |
-| Package upgrade | `.claude/skills/package-upgrade/` |
-
-### Code Quality
-| Trigger Keywords | Skill Folder |
-|:---|:---|
-| Code review | `.claude/skills/code-review/` |
-| Dual pass review | `.claude/skills/dual-pass-review/` |
-| Tasks code review | `.claude/skills/tasks-code-review/` |
-
-### Research & Problem Solving
-| Trigger Keywords | Skill Folder |
-|:---|:---|
-| Research | `.claude/skills/research/` |
-| Problem solving | `.claude/skills/problem-solving/` |
-| Sequential thinking | `.claude/skills/sequential-thinking/` |
-| Feature investigation | `.claude/skills/feature-investigation/` |
-| Feature implementation | `.claude/skills/feature-implementation/` |
-
-### Specialized
-| Trigger Keywords | Skill Folder |
-|:---|:---|
-| Better Auth | `.claude/skills/better-auth/` |
-| Shopify | `.claude/skills/shopify/` |
-| Payment integration | `.claude/skills/payment-integration/` |
-| Mobile development | `.claude/skills/mobile-development/` |
-| Web frameworks | `.claude/skills/web-frameworks/` |
-| Google ADK Python | `.claude/skills/google-adk-python/` |
-| Claude Code | `.claude/skills/claude-code/` |
-| Domain brainstormer | `.claude/skills/domain-name-brainstormer/` |
-| Branch comparison | `.claude/skills/branch-comparison/` |
-| Developer growth | `.claude/skills/developer-growth-analysis/` |
-| Docs seeker | `.claude/skills/docs-seeker/` |
-| Markdown novel viewer | `.claude/skills/markdown-novel-viewer/` |
-| Skill creator | `.claude/skills/skill-creator/` |
-
----
-
-## Usage Rules
-
-1. **Skills = Expert knowledge**: Load to get deep domain expertise.
-2. **Not always needed**: Simple code/fix tasks often don't need skills.
-3. **Read SKILL.md first**: It contains the main knowledge and may reference other files.
-4. **Combine with Agent + Command**: Skill enhances capability, doesn't replace workflow.

package/templates/router/workflows-guide.md

@@ -1,68 +0,0 @@
-# Workflows Guide
-
-> **Purpose**: Workflows define orchestration patterns for large, multi-step tasks.
-> **Location**: `.claude/workflows/<workflow-name>.md`
-> **When to load**: Only for complex tasks requiring multiple agents or phases.
-
----
-
-## Available Workflows
-
-| Trigger | Workflow | Load File | Use Case |
-|:---|:---|:---|:---|
-| New feature, large task | **Primary Workflow** | `.claude/workflows/primary-workflow.md` | Full feature development cycle |
-| Multi-agent coordination | **Orchestration Protocol** | `.claude/workflows/orchestration-protocol.md` | Sequential/parallel agent chains |
-| Dev rules, conventions | **Development Rules** | `.claude/workflows/development-rules.md` | Coding standards reference |
-| Documentation management | **Doc Management** | `.claude/workflows/documentation-management.md` | Docs lifecycle |
-
----
-
-## When to Use Workflows
-
-### Use Primary Workflow when:
-- Building a new feature from scratch
-- Task has multiple phases (plan → code → test → review)
-- Need to coordinate between planning, implementation, and testing
-
-### Use Orchestration Protocol when:
-- Multiple agents need to work together
-- Tasks can be parallelized
-- Need sequential handoff between agents
-
-### Skip Workflows when:
-- Simple bug fix
-- Single-file change
-- Quick question/answer
-- Small refactor
-
----
-
-## Workflow Patterns
-
-### Sequential Chaining
-```
-Planner → Developer → Tester → Reviewer
-```
-- Each agent completes before next begins
-- Pass context between agents
-- Use for dependent tasks
-
-### Parallel Execution
-```
-┌─ Developer (Feature A)
-│
-├─ Developer (Feature B)
-│
-└─ Tester (Write tests)
-```
-- Independent tasks run simultaneously
-- No file conflicts allowed
-- Plan merge strategy first
-
----
-
-## Usage Rules
-
-1. **Workflows = Orchestration**: For coordinating complex multi-step work.
-2. **Not for simple tasks**: Skip if task is straightforward.
-3. **Combine all layers**: Workflow coordinates Agents, who follow Commands, enhanced by Skills.

package/templates/workflows/README.md

@@ -1,241 +0,0 @@
-# 🔄 Workflows - Quy Trình Phối Hợp Nhiều Bước
-
-## Workflows là gì?
-
-**Workflows** là các "kịch bản phối hợp" cho những công việc lớn, phức tạp, cần nhiều bước và nhiều vai trò khác nhau. Giống như đạo diễn một bộ phim - cần kịch bản chi tiết ai làm gì, khi nào, theo thứ tự nào.
-
-**Khi nào cần Workflow:**
-- Tính năng mới lớn (cần plan → code → test → review → docs)
-- Nhiều người/vai trò phối hợp
-- Công việc kéo dài nhiều ngày/tuần
-- Cần checkpoints kiểm tra chất lượng
-
----
-
-## Danh Sách Workflows
-
-| Workflow | Dùng Khi Nào | Mô Tả |
-|----------|--------------|-------|
-| **primary-workflow.md** | Làm tính năng mới | Quy trình chuẩn từ plan đến hoàn thành |
-| **orchestration-protocol.md** | Phối hợp nhiều agents | Cách chuyển giao công việc giữa các vai trò |
-| **development-rules.md** | Quy tắc phát triển | Các nguyên tắc bắt buộc khi code |
-| **documentation-management.md** | Quản lý tài liệu | Quy trình cập nhật docs |
-
----
-
-## Chi Tiết Từng Workflow
-
-### 1. primary-workflow.md - Quy Trình Làm Tính Năng
-
-**Dùng khi:** Xây dựng tính năng mới từ đầu đến cuối
-
-**Các giai đoạn:**
-
-```
-GIAI ĐOẠN 1: LẬP KẾ HOẠCH
-├── Phân tích yêu cầu
-├── Thiết kế giải pháp
-├── Ước lượng thời gian
-└── Xác định rủi ro
-
-    ↓ Checkpoint: Kế hoạch được duyệt
-
-GIAI ĐOẠN 2: IMPLEMENT
-├── Viết code theo kế hoạch
-├── Tự review code
-├── Viết unit tests
-└── Fix bugs phát sinh
-
-    ↓ Checkpoint: Code pass tests
-
-GIAI ĐOẠN 3: KIỂM THỬ
-├── Chạy toàn bộ tests
-├── Test integration
-├── Test edge cases
-└── Fix issues
-
-    ↓ Checkpoint: Tất cả tests pass
-
-GIAI ĐOẠN 4: REVIEW
-├── Code review
-├── Security review
-├── Performance check
-└── Fix feedback
-
-    ↓ Checkpoint: Review approved
-
-GIAI ĐOẠN 5: HOÀN THÀNH
-├── Cập nhật docs
-├── Update changelog
-├── Merge code
-└── Deploy (nếu cần)
-```
-
----
-
-### 2. orchestration-protocol.md - Quy Trình Phối Hợp
-
-**Dùng khi:** Cần nhiều "chuyên gia" (agents) làm việc cùng nhau
-
-**Nguyên tắc chuyển giao:**
-
-```
-Agent A hoàn thành
-       ↓
-┌──────────────────────────┐
-│ CHUYỂN GIAO              │
-│ • Tóm tắt đã làm gì      │
-│ • Còn gì cần làm         │
-│ • Files đã thay đổi      │
-│ • Vấn đề cần lưu ý       │
-└──────────────────────────┘
-       ↓
-Agent B tiếp nhận
-```
-
-**Ví dụ phối hợp:**
-
-| Thứ tự | Agent | Làm gì | Chuyển cho |
-|--------|-------|--------|------------|
-| 1 | Planner | Lập kế hoạch | Developer |
-| 2 | Developer | Viết code | Tester |
-| 3 | Tester | Viết tests | Reviewer |
-| 4 | Reviewer | Review code | Docs Manager |
-| 5 | Docs Manager | Viết docs | (Hoàn thành) |
-
----
-
-### 3. development-rules.md - Quy Tắc Phát Triển
-
-**Dùng khi:** Luôn áp dụng khi viết code
-
-**Các quy tắc chính:**
-
-| Quy tắc | Ý nghĩa |
-|---------|---------|
-| **Minimal changes** | Chỉ sửa những gì cần thiết, không sửa thừa |
-| **Reuse before write** | Dùng code có sẵn trước, không viết lại |
-| **Test required** | Code mới phải có test |
-| **No magic numbers** | Không dùng số trực tiếp, dùng constants |
-| **Clear naming** | Đặt tên rõ ràng, dễ hiểu |
-| **File < 300 lines** | File quá dài thì phải tách |
-
----
-
-### 4. documentation-management.md - Quản Lý Tài Liệu
-
-**Dùng khi:** Cần cập nhật tài liệu dự án
-
-**Quy trình cập nhật:**
-
-```
-SAU KHI HOÀN THÀNH CÔNG VIỆC
-           ↓
-┌─────────────────────────────────┐
-│ Kiểm tra cần update docs nào?   │
-│ □ README.md (nếu thay đổi lớn)  │
-│ □ CHANGELOG.md (mọi thay đổi)   │
-│ □ docs/structure.md (cấu trúc)  │
-│ □ API docs (nếu thay đổi API)   │
-└─────────────────────────────────┘
-           ↓
-    Cập nhật các file cần thiết
-           ↓
-    Review và commit
-```
-
----
-
-## Ví Dụ Thực Tế
-
-### Ví dụ: Thêm Tính Năng "Gửi Email Thông Báo"
-
-**Bước 1: Load workflow**
-```
-AI load: primary-workflow.md
-```
-
-**Bước 2: Giai đoạn Planning**
-```
-Planner agent làm việc:
-- Phân tích: Cần gửi email khi nào?
-- Thiết kế: Dùng service nào? (SendGrid, SES?)
-- Ước lượng: ~2 ngày
-- Rủi ro: Rate limiting, spam filter
-
-→ Output: Kế hoạch chi tiết
-```
-
-**Bước 3: Giai đoạn Implement**
-```
-Developer agent làm việc:
-- Tạo EmailService class
-- Implement sendNotification()
-- Viết unit tests
-- Tích hợp vào notification flow
-
-→ Output: Code hoàn chỉnh + tests
-```
-
-**Bước 4: Giai đoạn Test**
-```
-Tester agent làm việc:
-- Chạy unit tests
-- Test gửi email thật (sandbox)
-- Test edge cases (email invalid, service down)
-
-→ Output: Test report
-```
-
-**Bước 5: Giai đoạn Review**
-```
-Reviewer agent làm việc:
-- Review code quality
-- Check security (no hardcoded credentials)
-- Check performance
-
-→ Output: Approval hoặc feedback
-```
-
-**Bước 6: Giai đoạn Docs**
-```
-Docs Manager agent làm việc:
-- Update README (new feature)
-- Update CHANGELOG
-- Add API documentation
-
-→ Output: Docs updated
-```
-
----
-
-## Khi Nào Cần Workflow?
-
-| Tình huống | Cần workflow? |
-|------------|---------------|
-| Sửa typo | ❌ Không |
-| Thêm button đơn giản | ❌ Không |
-| Thêm tính năng lớn | ✅ primary-workflow |
-| Refactor hệ thống | ✅ primary-workflow |
-| Nhiều người cùng làm | ✅ orchestration-protocol |
-| Cần docs chuẩn | ✅ documentation-management |
-
----
-
-## Tóm Tắt
-
-| Khái niệm | Giải thích |
-|-----------|------------|
-| **Workflow là gì** | Kịch bản phối hợp cho công việc lớn |
-| **Có bao nhiêu** | 4 workflows chính |
-| **Khi nào dùng** | Công việc lớn, nhiều bước, nhiều vai trò |
-| **Ai quyết định** | AI tự động khi nhận diện công việc phức tạp |
-
----
-
-## Xem Thêm
-
-- [Danh sách Agents (Vai trò)](../agents/README.md)
-- [Danh sách Commands (Quy trình)](../commands/README.md)
-- [Danh sách Skills (Kiến thức)](../skills/README.md)
-- [Router (Bộ định tuyến)](../router/README.md)

package/templates/workflows/orchestration-protocol.md

@@ -1,16 +0,0 @@
-# Orchestration Protocol
-
-#### Sequential Chaining
-Chain subagents when tasks have dependencies or require outputs from previous steps:
-- **Planning → Implementation → Testing → Review**: Use for feature development
-- **Research → Design → Code → Documentation**: Use for new system components
-- Each agent completes fully before the next begins
-- Pass context and outputs between agents in the chain
-
-#### Parallel Execution
-Spawn multiple subagents simultaneously for independent tasks:
-- **Code + Tests + Docs**: When implementing separate, non-conflicting components
-- **Multiple Feature Branches**: Different agents working on isolated features
-- **Cross-platform Development**: iOS and Android specific implementations
-- **Careful Coordination**: Ensure no file conflicts or shared resource contention
-- **Merge Strategy**: Plan integration points before parallel execution begins