antigravity-ide 4.2.1 → 4.3.3

package/.agent/workflows/create.md

@@ -1,59 +1,51 @@
 ---
-description: Muốn tạo tính năng mới từ A-Z? Dùng cái này.
+description: Muốn tạo tính năng mới hoặc dự án từ A-Z? Sử dụng bộ máy nhân sự chuyên nghiệp.
 ---
 
-# /create - Create Application
+# /create - Full-Cycle Product Creation
 
 $ARGUMENTS
 
 ---
 
-## Task
-
-This command starts a new application creation process.
-
-### Steps:
-
-1. **Request Analysis**
-   - Understand what the user wants
-   - If information is missing, use `conversation-manager` skill to ask
-
-2. **Project Planning**
-   - Use `project-planner` agent for task breakdown
-   - Determine tech stack
-   - Plan file structure
-   - Create plan file and proceed to building
-
-3. **Application Building (After Approval)**
-   - Orchestrate with `app-builder` skill
-   - Coordinate expert agents:
-     - `database-architect` → Schema
-     - `backend-specialist` → API
-     - `frontend-specialist` → UI
-
-4. **Preview**
-   - Start with `auto_preview.py` when complete
-   - Present URL to user
+## 🟢 PHASE 1: Requirements Discovery
+**Agent**: `product-manager` & `explorer-agent`
+**Mission**: Define the "What" and "Why."
+- **Action**: Analyze User Intent and map to project scale.
+- **Verification**: Ensure no conflict with existing system architecture.
+
+## 🟡 PHASE 2: Strategic Architecture (PLAN)
+**Agent**: `project-planner`
+**Mission**: Define the "How."
+- **Workflow Link**: Invokes `/plan` automatically.
+- **Artifact**: Create `docs/PLAN-{slug}.md`.
+- **Mandatory**: User MUST approve the plan before Phase 3 begins.
+
+## 🔵 PHASE 3: Surgical Execution
+**Agent**: `orchestrator`
+**Mission**: Direct the "Heavy Lifters."
+- **Choreography**: 
+  1. `database-architect` (Schema first).
+  2. `backend-specialist` (APIs/Logic).
+  3. `frontend-specialist` (Design/UI).
+  4. `documentation-writer` (Sync docs).
+
+## 🔴 PHASE 4: Professional Audit & Handoff
+**Agent**: `quality-inspector` & `test-engineer`
+**Mission**: Defend the Quality Gate.
+- **Verification**: Run `/test` and `/security` workflows.
+- **Artifact**: Create the final `walkthrough.md` with proof of work.
 
 ---
 
-## Usage Examples
-
-```
-/create blog site
-/create e-commerce app with product listing and cart
-/create todo app
-/create Instagram clone
-/create crm system with customer management
-```
+## Rules of Engagement:
+- **No Placeholders**: Every generated asset must be functional.
+- **Security-First**: No hardcoded keys, ever.
+- **Wow Factor**: Frontend work must use `ui-ux-pro-max` standards.
 
 ---
 
-## Before Starting
-
-If request is unclear, ask these questions:
-- What type of application?
-- What are the basic features?
-- Who will use it?
-
-Use defaults, add details later.
+## Examples:
+- `/create e-commerce dashboard with analytics`
+- `/create inventory management CLI in Rust`
+- `/create portfolio site with video backgrounds`

package/.agent/workflows/debug.md

@@ -1,103 +1,64 @@
 ---
-description: Gặp lỗi khó sửa? Để AI soi log và sửa giúp bạn.
+description: Gặp lỗi khó sửa? Để AI soi log và sửa giúp bạn theo quy trình chuyên nghiệp.
 ---
 
-# /debug - Systematic Problem Investigation
+# /debug - Systematic Debugging System
 
 $ARGUMENTS
 
 ---
 
-## Purpose
-
-This command activates DEBUG mode for systematic investigation of issues, errors, or unexpected behavior.
+## 🟢 PHASE 1: Forensic Discovery (The "Crime Scene")
+**Agent**: `explorer-agent` & `debugger`
+**Mission**: Isolate the exact point of failure.
+- **Action**: Read the Stack Trace or Terminal Logs.
+- **Action**: Locate the failing file:line.
+- **DNA Link**: Consult `rules/error-logging.md` to see if this is a known recurring issue.
+
+## 🟡 PHASE 2: Root Cause Analysis (RCA)
+**Agent**: `debugger`
+**Mission**: Formulate a hypothesis (Why is it broken?).
+- **Hypothesis Checklist**:
+  - Null/Undefined safety?
+  - Race condition / Sync error?
+  - Missing environment variable?
+  - Breaking dependency change?
+- **Artifact**: Propose the fix to the User with [Pros/Cons].
+
+## 🔵 PHASE 3: Surgical Repair
+**Agent**: `backend-specialist` or `frontend-specialist`
+**Mission**: Apply the targetted fix.
+- **Correction**: Wrap sensitive logic in `try...catch`.
+- **Defensive API**: Use Optional Chaining (`?.`) and Nullish Coalescing (`??`).
+
+## 🔴 PHASE 4: Verification & Post-Mortem
+**Agent**: `test-engineer` & `quality-inspector`
+**Mission**: Ensure the "Bleeding" has stopped.
+- **Action**: Run the failing test case to confirm FIX.
+- **Reporting**: Log the incident in `ERRORS.md`.
+- **Handoff**: Create a `walkthrough.md` explaining the fix.
 
 ---
 
-## Behavior
-
-When `/debug` is triggered:
-
-1. **Gather information**
-   - Error message
-   - Reproduction steps
-   - Expected vs actual behavior
-   - Recent changes
-
-2. **Form hypotheses**
-   - List possible causes
-   - Order by likelihood
-
-3. **Investigate systematically**
-   - Test each hypothesis
-   - Check logs, data flow
-   - Use elimination method
-
-4. **Fix and prevent**
-   - Apply fix
-   - Explain root cause
-   - Add prevention measures
-
----
-
-## Output Format
-
+## Output Format:
 ```markdown
-## 🔍 Debug: [Issue]
-
-### 1. Symptom
-[What's happening]
-
-### 2. Information Gathered
-- Error: `[error message]`
-- File: `[filepath]`
-- Line: [line number]
-
-### 3. Hypotheses
-1. ❓ [Most likely cause]
-2. ❓ [Second possibility]
-3. ❓ [Less likely cause]
+## 🐞 Debug Report: [Bug Title]
 
-### 4. Investigation
+### Root Cause
+[One sentence explanation]
 
-**Testing hypothesis 1:**
-[What I checked] → [Result]
+### The Fix
+[Diff or explanation]
 
-**Testing hypothesis 2:**
-[What I checked] → [Result]
-
-### 5. Root Cause
-🎯 **[Explanation of why this happened]**
-
-### 6. Fix
-```[language]
-// Before
-[broken code]
-
-// After
-[fixed code]
-```
-
-### 7. Prevention
-🛡️ [How to prevent this in the future]
-```
-
----
-
-## Examples
-
-```
-/debug login not working
-/debug API returns 500
-/debug form doesn't submit
-/debug data not saving
+### Verification
+- [ ] Test case passed
+- [ ] No regression found
+- [ ] Error logged in ERRORS.md
 ```
 
 ---
 
-## Key Principles
-
-- **Ask before assuming** - get full error context
-- **Test hypotheses** - don't guess randomly
-- **Explain why** - not just what to fix
-- **Prevent recurrence** - add tests, validation
+## Key Principles:
+- **Evidence-First**: Don't guess, use the logs.
+- **Regression-Aware**: Every fix must come with a test case to prevent it from returning.
+- **Clean Fix**: Don't use "band-aid" fixes unless it's a critical production outage.

package/.agent/workflows/deploy.md

@@ -2,175 +2,50 @@
 description: Code xong rồi? Đẩy lên Server/Vercel thôi.
 ---
 
-# /deploy - Production Deployment
+# /deploy - Production Release System
 
 $ARGUMENTS
 
 ---
 
-## Purpose
+## 🟢 PHASE 1: Pre-Flight Integrity Check
+**Agent**: `devops-engineer` & `security-auditor`
+**Mission**: Ensure the code is "Deployable."
+- **Action**: Run `npm run build` and `npm test`.
+- **Checklist**:
+  - [ ] Zero Lint/TS errors.
+  - [ ] All security patches applied (`npm audit`).
+  - [ ] Environment variables configured in Production.
 
-This command handles production deployment with pre-flight checks, deployment execution, and verification.
+## 🟡 PHASE 2: Artifact Creation & Gating
+**Agent**: `devops-engineer`
+**Mission**: Build the release.
+- **Action**: Create the production bundle/container.
+- **Gate**: Stop and verify the "Pre-deploy Checklist" success.
 
----
-
-## Sub-commands
+## 🔵 PHASE 3: Surgical Deployment
+**Agent**: `devops-engineer`
+**Mission**: Switch the traffic.
+- **Action**: Execute `vercel --prod`, `railway up`, or target platform command.
+- **Strategy**: Use Blue/Green or Canary if supported by the infrastructure.
 
-```
-/deploy            - Interactive deployment wizard
-/deploy check      - Run pre-deployment checks only
-/deploy preview    - Deploy to preview/staging
-/deploy production - Deploy to production
-/deploy rollback   - Rollback to previous version
-```
+## 🔴 PHASE 4: Health Audit & Verification
+**Agent**: `performance-optimizer` & `incident-responder`
+**Mission**: Verify the "Live" state.
+- **Action**: Check Production URL response time and status.
+- **Action**: Verify Database connectivity in the live environment.
+- **Artifact**: Provide the Production URL and Health Badge to the User.
 
 ---
 
-## Pre-Deployment Checklist
-
-Before any deployment:
-
-```markdown
-## 🚀 Pre-Deploy Checklist
-
-### Code Quality
-- [ ] No TypeScript errors (`npx tsc --noEmit`)
-- [ ] ESLint passing (`npx eslint .`)
-- [ ] All tests passing (`npm test`)
-
-### Security
-- [ ] No hardcoded secrets
-- [ ] Environment variables documented
-- [ ] Dependencies audited (`npm audit`)
-
-### Performance
-- [ ] Bundle size acceptable
-- [ ] No console.log statements
-- [ ] Images optimized
-
-### Documentation
-- [ ] README updated
-- [ ] CHANGELOG updated
-- [ ] API docs current
-
-### Ready to deploy? (y/n)
-```
+## Deployment Rules:
+- **No Yolo-Deploy**: Never deploy without a successful local build.
+- **Rollback Ready**: Always have a plan to return to the previous version.
+- **Zero Downtime**: Prioritize strategies that don't break the user experience.
 
 ---
 
-## Deployment Flow
-
-```
-┌─────────────────┐
-│  /deploy        │
-└────────┬────────┘
-         │
-         ▼
-┌─────────────────┐
-│  Pre-flight     │
-│  checks         │
-└────────┬────────┘
-         │
-    Pass? ──No──► Fix issues
-         │
-        Yes
-         │
-         ▼
-┌─────────────────┐
-│  Build          │
-│  application    │
-└────────┬────────┘
-         │
-         ▼
-┌─────────────────┐
-│  Deploy to      │
-│  platform       │
-└────────┬────────┘
-         │
-         ▼
-┌─────────────────┐
-│  Health check   │
-│  & verify       │
-└────────┬────────┘
-         │
-         ▼
-┌─────────────────┐
-│  ✅ Complete    │
-└─────────────────┘
-```
-
----
-
-## Output Format
-
-### Successful Deploy
-
-```markdown
-## 🚀 Deployment Complete
-
-### Summary
-- **Version:** v1.2.3
-- **Environment:** production
-- **Duration:** 47 seconds
-- **Platform:** Vercel
-
-### URLs
-- 🌐 Production: https://app.example.com
-- 📊 Dashboard: https://vercel.com/project
-
-### What Changed
-- Added user profile feature
-- Fixed login bug
-- Updated dependencies
-
-### Health Check
-✅ API responding (200 OK)
-✅ Database connected
-✅ All services healthy
-```
-
-### Failed Deploy
-
-```markdown
-## ❌ Deployment Failed
-
-### Error
-Build failed at step: TypeScript compilation
-
-### Details
-```
-error TS2345: Argument of type 'string' is not assignable...
-```
-
-### Resolution
-1. Fix TypeScript error in `src/services/user.ts:45`
-2. Run `npm run build` locally to verify
-3. Try `/deploy` again
-
-### Rollback Available
-Previous version (v1.2.2) is still active.
-Run `/deploy rollback` if needed.
-```
-
----
-
-## Platform Support
-
-| Platform | Command | Notes |
-|----------|---------|-------|
-| Vercel | `vercel --prod` | Auto-detected for Next.js |
-| Railway | `railway up` | Needs Railway CLI |
-| Fly.io | `fly deploy` | Needs flyctl |
-| Docker | `docker compose up -d` | For self-hosted |
-
----
-
-## Examples
-
-```
-/deploy
-/deploy check
-/deploy preview
-/deploy production --skip-tests
-/deploy rollback
-```
+## Examples:
+- `/deploy production`
+- `/deploy preview`
+- `/deploy check`

package/.agent/workflows/document.md

@@ -1,27 +1,50 @@
 ---
-description: Lười viết docs? Để AI tự viết cho.
+description: Lười viết docs? Để AI tự viết cho chuyên nghiệp và đầy đủ.
 ---
 
-# /document - Auto Documentation
+# /document - Knowledge & Documentation System
 
 $ARGUMENTS
 
 ---
 
-## Task
-Keep documentation in sync with code.
+## 🟢 PHASE 1: Codebase Extraction
+**Agent**: `explorer-agent` & `documentation-writer`
+**Mission**: Understand what needs explaining.
+- **Action**: Extract Docstrings, Type Definitions, and API routes.
+- **Action**: Identify "Knowledge Gaps" in current documentation.
 
-### Steps:
-1.  **Scan Codebase**: Read function signatures and comments.
-2.  **Generate API Docs**: Create Swagger/OpenAPI spec.
-3.  **Update README**: Refresh feature list and usage examples.
-4.  **Internal Docs**: Update `ARCHITECTURE.md` if structure changed.
+## 🟡 PHASE 2: Logic & Structural Drafting
+**Agent**: `documentation-writer`
+**Mission**: Create the narrative.
+- **Protocol**: 
+  - Why does this code exist? (Business Logic).
+  - How do I use it? (Quick Start/Examples).
+  - What are the risks? (Caveats/Notes).
+- **Format**: Use GitHub Flavored Markdown (GFM) with appropriate Alerts.
+
+## 🔵 PHASE 3: Surgical Update
+**Agent**: `documentation-writer`
+**Mission**: Apply the knowledge.
+- **Action**: Create or Update `README.md`, `API.md`, or code comments.
+- **Standard**: Follow [documentation-templates](file:///skills/documentation-templates/SKILL.md).
+
+## 🔴 PHASE 4: Clarity Audit & Sync
+**Agent**: `quality-inspector` & `seo-specialist`
+**Mission**: Ensure the docs are discoverable and clear.
+- **Verification**: Run `textlint` and check for broken links.
+- **Artifact**: Create a unified `walkthrough.md` if this was a major doc update.
+
+---
+
+## Documentation Mandates:
+- **No Stale Docs**: Documentation must match current code state.
+- **AI-Ready**: Use structured headers to help other agents find context.
+- **User-Centric**: Write for the persona appropriate for the file (Dev vs User).
 
 ---
 
-## Usage
-```
-/document api      # Generate API reference
-/document readme   # Update README.md
-/document all      # Full documentation refresh
-```
+## Examples:
+- `/document this entire directory`
+- `/document the authentication flow`
+- `/document generate API reference`

package/.agent/workflows/enhance.md

@@ -2,62 +2,47 @@
 description: Muốn sửa màu, thêm nút, sửa logic nhỏ? Vào đây.
 ---
 
-# /enhance - Update Application
+# /enhance - Incremental Feature Upgrade
 
 $ARGUMENTS
 
 ---
 
-## Task
-
-This command adds features or makes updates to existing application.
-
-### Steps:
-
-1. **Understand Current State**
-   - Load project state with `python .agent/scripts/session_manager.py info`
-   - Understand existing features, tech stack
-
-2. **Plan Changes**
-   - Determine what will be added/changed
-   - Detect affected files
-   - Check dependencies
-
-3. **Present Plan to User** (for major changes)
-   ```
-   "To add admin panel:
-   - I'll create 15 new files
-   - Update 8 files
-   - Takes ~10 minutes
-   
-   Should I start?"
-   ```
-
-4. **Apply**
-   - Call relevant agents
-   - Make changes
-   - Test
-
-5. **Update Preview**
-   - Hot reload or restart
+## 🟢 PHASE 1: Impact Discovery
+**Agent**: `explorer-agent` & `frontend-specialist`
+**Mission**: Map the "Blast Radius."
+- **Action**: Identify all components and services affected by the change.
+- **Action**: Check for breaking CSS or Logic changes.
+
+## 🟡 PHASE 2: Enhancement Plan
+**Agent**: `project-planner`
+**Mission**: Strategic addition.
+- **Action**: Create a mini-plan in the `task.md`.
+- **Note**: For major UI overhauls, invoke `/brainstorm` first.
+
+## 🔵 PHASE 3: Surgical Implementation
+**Agent**: `frontend-specialist` or `backend-specialist`
+**Mission**: Add the "WOW" factor.
+- **Action**: Implement changes using `ui-ux-pro-max` standards.
+- **Standard**: Follow [clean-code](file:///skills/clean-code/SKILL.md) and [tailwind-patterns](file:///skills/tailwind-patterns/SKILL.md).
+
+## 🔴 PHASE 4: Professional Verification
+**Agent**: `quality-inspector` & `test-engineer`
+**Mission**: Verify the delta.
+- **Action**: Run `/preview` to playtest the UI.
+- **Action**: Verify that existing tests still pass (Regression Check).
+- **Handoff**: Document the change in `walkthrough.md`.
 
 ---
 
-## Usage Examples
-
-```
-/enhance add dark mode
-/enhance build admin panel
-/enhance integrate payment system
-/enhance add search feature
-/enhance edit profile page
-/enhance make responsive
-```
+## Enhancement Principles:
+- **Non-Destructive**: Don't break existing features to add new ones.
+- **Consistent UI**: New elements must match the existing Design System.
+- **Atomic Commits**: Group project changes in logical git commits.
 
 ---
 
-## Caution
-
-- Get approval for major changes
-- Warn on conflicting requests (e.g., "use Firebase" when project uses PostgreSQL)
-- Commit each change with git
+## Examples:
+- `/enhance add dark mode toggle`
+- `/enhance integrate Google Auth`
+- `/enhance make the landing page sections fade in`

package/.agent/workflows/monitor.md

@@ -1,26 +1,47 @@
 ---
-description: Server có ổn không? Cài giám sát ngay.
+description: Server và Pipeline có ổn không? Cài đặt giám sát ngay.
 ---
 
-# /monitor - DevOps & Observability
+# /monitor - Full-Stack Observability & Health
 
 $ARGUMENTS
 
 ---
 
-## Task
-Setup logging and monitoring for Production.
+## 🟢 PHASE 1: Topology Discovery
+**Agent**: `devops-engineer` & `explorer-agent`
+**Mission**: Map the "Heartbeat" locations.
+- **Action**: Identify critical services, databases, and CI/CD pipelines.
+- **Action**: Locate existing logging and metrics configuration.
 
-### Steps:
-1.  **Logging**: Config structured logging (JSON format).
-2.  **Error Tracking**: Setup Sentry/LogRocket/Datadog.
-3.  **Health Checks**: Create `/health` API endpoint.
-4.  **Alerting**: Define thresholds for CPU/Memory warnings.
+## 🟡 PHASE 2: Sourcing & Alerting Logic
+**Agent**: `incident-responder` or `devops-engineer`
+**Mission**: Define "Healthy."
+- **Action**: Set thresholds for Error Rates, Latency (p99), and Resource Saturation.
+- **Action**: Configure Alerting routes (Slack, Email, PagerTree).
+
+## 🔵 PHASE 3: Surgical Instrumentation
+**Agent**: `devops-engineer`
+**Mission**: Install the "Nervous System."
+- **Action**: Setup OpenTelemetry, Prometheus, or Grafana Dashboards.
+- **Action**: Implement log rotation and sanitization filters.
+
+## 🔴 PHASE 4: Reliability Audit
+**Agent**: `quality-inspector`
+**Mission**: Verify the "Watchman."
+- **Verification**: Trigger a mock failure to test alert delivery.
+- **Reporting**: Report the Dashboard URL or "Health Status" to the User.
+
+---
+
+## Observability Principles:
+- **Zero Blindspots**: Monitor the "Critical Path" end-to-end.
+- **No Noise**: Alerts must be actionable; suppress low-signal chatter.
+- **Evidence-Based**: Operational decisions must be based on metrics, not intuition.
 
 ---
 
-## Usage
-```
-/monitor setup sentry   # Install Sentry SDK
-/monitor check          # Verify observability
-```
+## Examples:
+- `/monitor setup uptime tracking`
+- `/monitor dashboard for backend latency`
+- `/monitor check pipeline reliability`