anchi-kit 1.1.2

package/docs/AI_CONTRACT.md

@@ -0,0 +1,150 @@
+# AI Execution Contract
+
+> **Hiến pháp AI** - Luật bắt buộc mà AI phải tuân thủ
+
+---
+
+## 📜 Rule Priority (Thứ Tự Ưu Tiên)
+
+```
+1. SAFETY RULES     ← Không bao giờ vi phạm
+2. PROJECT RULES    ← Từ preset + project config
+3. COMMAND RULES    ← Từ command đang thực thi
+4. USER REQUEST     ← Yêu cầu hiện tại
+```
+
+> **Nếu conflict: Rule cao hơn thắng.**
+
+---
+
+## 🔴 Level 1: Safety Rules (Bất khả xâm phạm)
+
+| Rule | Description |
+|------|-------------|
+| **S1** | Không xóa file quan trọng (.env, config, database) mà không hỏi |
+| **S2** | Không commit/push mà không có confirmation |
+| **S3** | Không chạy destructive commands (rm -rf, DROP TABLE) tự động |
+| **S4** | Không expose secrets trong code hoặc logs |
+| **S5** | Phải backup trước khi thay đổi lớn |
+
+**Violation Action:** DỪNG ngay, hỏi user.
+
+---
+
+## 🟡 Level 2: Project Rules (Từ Preset)
+
+### Rapid Preset
+```json
+{
+  "allowed": ["any structure", "SQLite", "quick patterns"],
+  "forbidden": [],
+  "checks": ["syntax only"]
+}
+```
+
+### Professional Preset
+```json
+{
+  "allowed": ["modular structure", "PostgreSQL"],
+  "forbidden": ["logic in components", "direct DB in routes"],
+  "checks": ["lint", "type-check", "test coverage > 60%"]
+}
+```
+
+### Enterprise Preset
+```json
+{
+  "allowed": ["DDD structure", "PostgreSQL + Redis"],
+  "forbidden": ["anemic domain", "circular deps", "skip reviews"],
+  "checks": ["lint", "type-check", "test coverage > 80%", "security scan"]
+}
+```
+
+**Violation Action:** Cảnh báo, đề xuất sửa, yêu cầu approval.
+
+---
+
+## 🟢 Level 3: Command Rules
+
+Mỗi command có rules riêng:
+
+| Command | Rules |
+|---------|-------|
+| `/do --fast` | Skip reviews OK, nhưng vẫn check Safety |
+| `/do --careful` | Extra validation, require tests |
+| `/commit` | Require confirmation trước push |
+| `/design` | Follow design system nếu có |
+| `/generate` | Match existing patterns |
+
+---
+
+## ⚖️ Conflict Resolution
+
+```
+User: /do "Add logic to Button component"
+AI: ⚠️ CONFLICT DETECTED
+
+    User Request: Add logic to Button
+    Project Rule: "logic in components forbidden" (Professional)
+    
+    Resolution:
+    → Create ButtonService for logic
+    → Button component calls service
+    
+    Proceed with corrected approach? [y/n]
+```
+
+---
+
+## 📋 Enforcement Checklist
+
+Trước mỗi action, AI PHẢI kiểm tra:
+
+- [ ] Safety rules OK?
+- [ ] Project rules OK?
+- [ ] Command rules OK?
+- [ ] Conflicts resolved?
+
+**Chỉ proceed khi tất cả OK hoặc đã có user approval.**
+
+---
+
+## 🔄 Violation Handling
+
+```
+┌─────────────────────────────────────────────────────────────┐
+│ VIOLATION DETECTED                                          │
+├─────────────────────────────────────────────────────────────┤
+│                                                             │
+│  Level 1 (Safety):   STOP immediately, ask user             │
+│  Level 2 (Project):  Warn, suggest fix, ask approval        │
+│  Level 3 (Command):  Adjust approach, inform user           │
+│                                                             │
+└─────────────────────────────────────────────────────────────┘
+```
+
+---
+
+## 📝 Audit Trail
+
+Mọi decision được log:
+
+```json
+{
+  "timestamp": "2024-12-26T10:30:00",
+  "command": "/do 'Add payment'",
+  "rules_checked": ["S1", "S2", "P3", "C1"],
+  "violations": [],
+  "decision": "Proceed with Stripe integration",
+  "approved_by": "auto"
+}
+```
+
+---
+
+## 🔗 Related Files
+
+- `CURSOR.md` - AI behavior configuration
+- `.cursorrules` - Project-specific rules
+- `.cursor/team.yaml` - Team roles
+- `presets/*/contract.json` - Preset contracts

package/docs/ALL_COMMANDS.md

@@ -0,0 +1,381 @@
+# 📖 anchi-kit Command Reference v2.0
+
+> **Philosophy:** "Less but Quality" - 100% Dynamic Orchestration, Zero Static Mappings
+
+---
+
+## 🖥️ Terminal Admin Commands (CLI)
+
+Các lệnh quản trị chạy trong **Terminal** (`npx anchi-kit ...`). Dùng để cài đặt, kiểm tra sức khỏe và quản lý "bộ não" AI.
+
+| Command                       | Alias      | Description                                                   |
+| :---------------------------- | :--------- | :------------------------------------------------------------ |
+| **`npx anchi-kit`**           | -          | **Smart Setup** (Auto-Detect IDE, Secure Install, Dependency) |
+| `npx anchi-kit brain`         | `b`        | Xem tổng quan Memory, Learning & Audit 🧠                     |
+| `brain memory`                | `b memory` | Quản lý ký ức (`show`, `add`, `search`, `clear`, `snapshot`)  |
+| `brain learn`                 | `b learn`  | Học đa dự án (`register`, `suggest`)                          |
+| `brain audit`                 | `b audit`  | Xem nhật ký AI (`show`, `summary`)                            |
+| `brain analyze <file>`        | -          | Phân tích ảnh hưởng khi sửa file                              |
+| **`npx anchi-kit doctor`**    | -          | Kiểm tra cấu hình hệ thống                                    |
+| **`npx anchi-kit clean`**     | -          | Dọn dẹp logs/memory (`--memory`, `--audit`, `--all`)          |
+| **`npx anchi-kit dashboard`** | **`dash`** | Mở Dashboard trên trình duyệt (Port 4567)                     |
+| **`npx anchi-kit tui`**       | -          | Giao diện Terminal tương tác                                  |
+| **`npx anchi-kit team`**      | -          | Quản lý Team & Sync Memory (Hive Mind)                        |
+| **`npx anchi-kit patterns`**  | **`pat`**  | Phát hiện code patterns trong dự án                           |
+| `npx anchi-kit update`        | `u`        | Kiểm tra bản cập nhật mới                                     |
+| `npx anchi-kit uninstall`     | -          | Gỡ bỏ toolkit khỏi dự án                                      |
+
+### CLI Flags
+
+| Flag            | Description                           |
+| --------------- | ------------------------------------- |
+| `-v, --version` | Hiển thị version (`anchi-kit v1.0.0`) |
+| `-y, --force`   | Bỏ qua prompts xác nhận               |
+| `--no-open`     | Dashboard: Không tự mở browser        |
+
+---
+
+## � Chat Agent Commands (IDE)
+
+Các lệnh gõ trực tiếp trong **cửa sổ Chat** của Cursor/Gemini. Dùng để thực thi công việc hàng ngày.
+
+### �💎 The Core Trinity (Must Know)
+
+These are the ONLY commands you need to memorize:
+
+| Command      | Function             | Dynamic Intelligence                                       |
+| ------------ | -------------------- | ---------------------------------------------------------- |
+| **`/start`** | Init/Resume Project  | Detects new vs existing → selects planner or scout         |
+| **`/do`**    | **Execute Anything** | 9-step orchestration, auto-selects agents & skills         |
+| **`/clean`** | Self-Healing ✨ NEW  | Smart cleanup based on flags (--memory, --audit, --config) |
+
+**Example:**
+
+```bash
+/start              # AI: "Welcome back! Last task: Login API (90% done)"
+/do "add payment"   # Auto-activates: payment-integration + security-auditor
+/clean --memory     # Auto-selects: docs-manager for memory management
+```
+
+---
+
+## ⚡ Power User Commands (5)
+
+For specific workflows:
+
+| Command            | Function             | Auto-Detection                                                             |
+| ------------------ | -------------------- | -------------------------------------------------------------------------- |
+| **`/demo`**        | Auto-preview         | Detects Docker/Vercel/K8s → chooses devops-engineer or fullstack-developer |
+| **`/skill-learn`** | Teach AI new tech    | researcher → skill-creator workflow                                        |
+| **`/why`**         | Explain AI decisions | Shows agent selection, skill activation, risk assessment                   |
+| **`/status`**      | Project health       | project-manager (--memory → + docs-manager)                                |
+| **`/memory`**      | Context management   | docs-manager (--save → + journal-writer)                                   |
+
+---
+
+## 🛠️ Development Commands (4)
+
+| Command          | Function                | Smart Routing                                                   |
+| ---------------- | ----------------------- | --------------------------------------------------------------- |
+| **`/generate`**  | Generate code           | Auto-detects: `*Page` → ui-ux-designer, `*API` → database-admin |
+| **`/scout`**     | Explore codebase        | scout (--external → scout-external + researcher)                |
+| **`/integrate`** | Third-party integration | Auto-escalates security for payment/auth                        |
+| **`/migrate`**   | Preset/DB migration     | ddd-architect (database → + database-admin)                     |
+
+**Example:**
+
+```bash
+/generate "LoginPage"    # Auto: ui-ux-designer + fullstack-developer
+/scout --external        # Auto: scout-external + MCP tools
+/integrate "Stripe"     # Auto: security-auditor escalation
+```
+
+---
+
+## 🧪 Quality Assurance Commands (3)
+
+| Command       | Function     | Mode-Based Agents                                         |
+| ------------- | ------------ | --------------------------------------------------------- |
+| **`/test`**   | Run tests    | tester (--e2e → + playwright-testing, --fix → + debugger) |
+| **`/review`** | Code review  | code-reviewer (--security → security-auditor)             |
+| **`/audit`**  | View AI logs | code-reviewer (security → security-auditor)               |
+
+**Example:**
+
+```bash
+/test --e2e          # Auto: tester + playwright-testing skill
+/review --security   # Auto: security-auditor escalation
+```
+
+---
+
+## 🎨 Design Commands (3)
+
+| Command       | Function         | Multi-Mode                                                  |
+| ------------- | ---------------- | ----------------------------------------------------------- |
+| **`/design`** | UI/Domain design | ui-ux-designer (--domain → ddd-architect, --3d → + threejs) |
+| **`/theme`**  | Theme management | design-system-architect (dark mode → + ui-ux-designer)      |
+| **`/docs`**   | Documentation    | docs-manager (api docs → + fullstack-developer)             |
+
+**Example:**
+
+```bash
+/design "Dashboard"              # Auto: ui-ux-designer
+/design "User entity" --domain   # Auto: ddd-architect
+/design "3D scene" --3d          # Auto: ui-ux-designer + threejs
+```
+
+---
+
+## 🔧 Version Control Commands (2)
+
+| Command       | Function         | Smart Detection                                                 |
+| ------------- | ---------------- | --------------------------------------------------------------- |
+| **`/commit`** | Git commit       | git-manager (BREAKING → + changelog, >20 files → code-reviewer) |
+| **`/undo`**   | Rollback changes | git-manager (workflow → + ddd-architect)                        |
+
+**Example:**
+
+```bash
+/commit "BREAKING: remove API v1"   # Auto: + changelog-generator
+/commit (23 files modified)         # Auto: code-reviewer warning
+```
+
+---
+
+## ⚙️ Utilities (3)
+
+| Command       | Function         | Conditional Logic                              |
+| ------------- | ---------------- | ---------------------------------------------- |
+| **`/config`** | View/edit config | project-manager (--validate → + code-reviewer) |
+| **`/help`**   | Show help        | docs-manager (specific → + researcher)         |
+
+---
+
+## 📊 All 22 Commands Summary
+
+### By Category
+
+**Essential (3):** /start, /do, /clean  
+**Power User (5):** /demo, /skill-learn, /why, /status, /memory  
+**Development (4):** /generate, /scout, /integrate, /migrate  
+**Quality (3):** /test, /review, /audit  
+**Design (3):** /design, /theme, /docs  
+**VCS (2):** /commit, /undo  
+**Utilities (2):** /config, /help
+
+> **Note:** /dashboard và /tui là các lệnh CLI (`npx anchi-kit dashboard/tui`), không phải Chat commands.
+
+---
+
+## 🎯 Command Decision Tree
+
+```
+Need to BUILD something?
+  → /do "what you want"
+
+Need to TEST?
+  → /test
+
+Need to REVIEW code?
+  → /review [--security|--performance]
+
+Need to DESIGN?
+  → /design "what to design" [--domain|--3d]
+
+Need to UNDERSTAND?
+  → /scout "what to find"
+  → /why (explain last decision)
+
+Need to MAINTAIN?
+  → /clean [--memory|--audit|--all]
+  → /status
+
+Need to LEARN new tech?
+  → /skill-learn "technology name"
+```
+
+---
+
+## 🚀 Dynamic Orchestration Examples
+
+### Example 1: Multi-Skill Activation
+
+```bash
+User: /do "create dashboard with 3D charts and real-time updates"
+
+AI Decision:
+✅ Detected "dashboard" → ui-ux-designer (primary)
+✅ Detected "3d" → threejs skill
+✅ Detected "charts" → d3-visualization skill
+✅ Detected "real-time" → websockets (auto)
+✅ Activated: aesthetic, frontend-development skills
+
+Agents: ui-ux-designer, fullstack-developer
+Skills: threejs, d3-visualization, ui-styling, websockets
+```
+
+### Example 2: Security Auto-Escalation
+
+```bash
+User: /do "update payment processing"
+
+AI Decision:
+✅ Detected "payment" → payment-integration skill
+✅ Auto-escalated to security-auditor (requires_security_review: true)
+✅ Activated: backend-development, databases skills
+✅ Risk level: HIGH (automatic review required)
+
+Agents: fullstack-developer, security-auditor
+Skills: payment-integration, security-best-practices, backend-development
+```
+
+### Example 3: Context-Based Agent Selection
+
+```bash
+# Scenario A: Docker project
+User: /demo
+Detection: docker-compose.yml exists
+Agent: devops-engineer
+Skills: devops, docker-management
+
+# Scenario B: Next.js project
+User: /demo
+Detection: package.json only
+Agent: fullstack-developer
+Skills: web-frameworks, frontend-development
+```
+
+---
+
+## 🧠 How AI Chooses
+
+**1. Analyze Request:**
+
+- Extract keywords ("payment", "3d", "test")
+- Identify file patterns (_.tsx, _.py)
+- Detect project structure (docker-compose, k8s)
+
+**2. Select Agents:**
+
+- Check agent capabilities
+- Apply conditionals (if "payment" → security-auditor)
+- Build execution chain
+
+**3. Activate Skills:**
+
+- Match keywords to skills
+- Check package.json for tools
+- Auto-activate dependencies
+
+**4. Execute with Intelligence:**
+
+- Follow orchestration.yaml chains
+- Apply failure-modes.yaml safety
+- Log to .ai-audit/ for transparency
+
+**5. Explain with /why:**
+
+- Show all decisions
+- Display activated skills
+- Present risk assessment
+
+---
+
+## 💡 Pro Tips
+
+### 1. Let AI Choose
+
+```bash
+❌ Don't: "use fullstack-developer to add login"
+✅ Do: /do "add login"
+# AI will choose fullstack-developer + security-auditor automatically
+```
+
+### 2. Use Flags for Control
+
+```bash
+/do "add payment" --careful      # More thorough
+/do "fix typo" --fast            # Quick mode
+/test --e2e                      # Specific test type
+```
+
+### 3. Understand with /why
+
+```bash
+/do "complex task"
+/why  # See why AI chose specific agents & skills
+```
+
+### 4. Trust the System
+
+```bash
+/do "add Stripe with mobile app and analytics"
+# AI will activate:
+# - payment-integration (Stripe)
+# - mobile-development (mobile app)
+# - csv-data-analyzer (analytics)
+# - security-auditor (auto-escalate)
+# All working together seamlessly!
+```
+
+---
+
+## 🗑️ Removed Commands
+
+These commands are now **subsumed by `/do`** with automatic routing:
+
+**Deprecated (don't use):**
+
+- `/plan`, `/plan:*` → Use `/do --plan-only`
+- `/cook`, `/cook:*` → Use `/do`
+- `/fix`, `/fix:*` → Use `/do "fix ..."`
+- `/auto:*` → Use `/do`
+- `/git:*` → Use `/commit` or `/undo`
+
+**Total consolidation:** 40+ commands → 22 commands (45% reduction)
+
+---
+
+## 📚 Learn More
+
+- **[Dynamic Orchestration Guide](ORCHESTRATION.md)** - How AI makes decisions
+- **[Agent & Skill Index](../.cursor/agent-skill-index.yaml)** - Complete mapping
+- **[Implementation Plan](../docs/)** - Technical details
+
+---
+
+## 🎓 Quick Start Guide
+
+**Day 1: Basics**
+
+```bash
+/start              # Begin
+/do "simple task"   # Build
+/test               # Verify
+/commit             # Ship
+```
+
+**Day 2: Power User**
+
+```bash
+/do "complex task" --careful
+/why                # Understand
+/scout "feature"    # Explore
+/clean              # Maintain
+```
+
+**Day 3: Advanced**
+
+```bash
+/skill-learn "new tech"
+/design --domain
+/integrate "external service"
+/review --security
+```
+
+---
+
+**v2.0 - 100% Dynamic Orchestration** 🚀
+
+**22 Commands | 55 Skills | 21 Agents | Zero Static Mappings**