Skip to main content

​
Slash Commands

Slash commands are structured workflow procedures that guide the AI through specific types of tasks. Antigravity Kit includes 11 specialized workflows that handle everything from brainstorming to deployment.

​
Available Commands

Antigravity Kit provides 11 slash command workflows:
CommandDescription
/brainstormExplore options before implementation
/createCreate new features or apps
/debugSystematic debugging
/deployDeploy application
/enhanceImprove existing code
/orchestrateMulti-agent coordination
/planCreate task breakdown
/previewPreview changes locally
/statusCheck project status
/testGenerate and run tests
/ui-ux-pro-maxDesign with 50 styles

​
When to Use Each Workflow

​
/brainstorm - Explore Before You Build

Best For

  • Unclear or vague requirements
  • Need to explore multiple options
  • Complex problems that need breaking down
  • Early-stage feature planning
How it works:
  1. Loads the brainstorming skill
  2. Applies Socratic questioning method
  3. Asks strategic questions to clarify requirements
  4. Outputs a structured discovery document
Example usage: 
/brainstorm authentication system
Expected output:
  • 3-5 strategic questions about your requirements
  • Analysis of different authentication approaches (JWT, OAuth, session-based)
  • Pros and cons of each option
  • Recommended approach with rationale
  • Implementation considerations
Always use /brainstorm before starting major features. The 5 minutes spent on discovery can save hours of rework.

​
/create - Build New Features

Best For

  • New features in existing projects
  • Small-to-medium complexity tasks
  • Single domain work (frontend OR backend)
  • Rapid prototyping
How it works:
  1. Detects project type (web/mobile/api/game)
  2. Loads app-builder skill + domain-specific skills
  3. Selects appropriate template if creating from scratch
  4. Generates project structure and code
  5. Runs quick validation (checklist.py)
Example usage: 
/create landing page with hero section
/create REST API for blog posts
/create user profile component
Expected output:
  • Complete code files for the feature
  • Proper project structure
  • Dependencies installed (if needed)
  • Basic tests included
  • Documentation on how to use the feature

​
/orchestrate - Multi-Agent Coordination

Best For

  • Full-stack features
  • Complex multi-step tasks
  • Projects requiring multiple specialist agents
  • Large features spanning multiple domains
How it works:
  1. Loads parallel-agents skill
  2. Decomposes task into subtasks
  3. Assigns each subtask to appropriate specialist agent
  4. Coordinates sequential execution with context switching
  5. Ensures code coherence across all parts
  6. Runs full verification suite
Example usage: 
/orchestrate Build a Next.js dashboard with authentication
Expected output: For a full-stack dashboard with authentication: Frontend (via @frontend-specialist):
  • app/dashboard/page.tsx - Server Component
  • components/DashboardLayout.tsx - Layout component
  • components/LoginForm.tsx - Login UI
  • lib/auth-client.ts - Client-side auth utilities
Backend (via @backend-specialist):
  • app/api/auth/login/route.ts - Login endpoint
  • app/api/auth/logout/route.ts - Logout endpoint
  • lib/jwt.ts - Token management
  • middleware.ts - Auth middleware
Database (via @database-architect):
  • prisma/schema.prisma - User and Session models
  • prisma/migrations/ - Database migrations
Testing (via @test-engineer):
  • tests/auth.spec.ts - Playwright E2E tests
  • tests/dashboard.spec.ts - Dashboard tests
The orchestrator maintains consistency across all code, ensuring imports, types, and API contracts all align perfectly.

​
/debug - Systematic Debugging

Best For

  • Bug reports and error messages
  • Unexpected application behavior
  • Performance issues
  • Failed test cases
How it works:
  1. Loads systematic-debugging skill
  2. Analyzes error logs and stack traces
  3. Applies root cause analysis methodology
  4. Suggests fixes with code examples
  5. Verifies the fix works
Example usage: 
/debug why login fails
/debug TypeError: Cannot read property 'name' of undefined
/debug page loads slowly
Expected output:
  • Root cause analysis of the issue
  • Step-by-step explanation of what went wrong
  • Code fix with before/after examples
  • Test to verify the fix
  • Prevention tips to avoid similar issues

​
/test - Generate and Run Tests

Best For

  • Need test coverage for new features
  • Before deploying to production
  • After major code changes
  • Continuous integration setup
How it works:
  1. Loads testing-patterns + webapp-testing skills
  2. Detects test framework (Jest, Vitest, Playwright)
  3. Generates appropriate test cases
  4. Executes test suite
  5. Reports coverage and results
Example usage: 
/test authentication flow
/test API endpoints
/test user registration
Expected output:
  • Unit tests for functions and components
  • Integration tests for API endpoints
  • E2E tests for critical user flows
  • Test execution results
  • Coverage report
Always run /test before deploying to production. It catches issues that might not be obvious during development.

​
/deploy - Deploy to Production

Best For

  • Ready to ship to production
  • After all tests pass
  • Need production URL for stakeholders
  • CI/CD pipeline setup
How it works:
  1. Loads deployment-procedures skill
  2. Detects target platform (Vercel, AWS, Docker, etc.)
  3. Prepares build artifacts
  4. Executes deployment scripts
  5. Runs health checks
  6. Outputs deployment URL
Example usage: 
/deploy to Vercel
/deploy using Docker
/deploy to AWS
Expected output:
  • Build process completion
  • Deployment configuration files
  • Deployment status updates
  • Production URL
  • Post-deployment health check results

​
/plan - Create Task Breakdown

Best For

  • Large projects requiring planning
  • Need time estimates for stakeholders
  • Team coordination and sprint planning
  • Complex features with dependencies
How it works:
  1. Loads plan-writing + architecture skills
  2. Analyzes requirements
  3. Breaks down into tasks with estimates
  4. Identifies dependencies between tasks
  5. Outputs structured plan with milestones
Example usage: 
/plan e-commerce checkout system
/plan migration to microservices
Expected output:
  • Phased breakdown of work
  • Task list with time estimates
  • Dependency graph
  • Risk assessment
  • Suggested milestones
  • Resource requirements

​
/enhance - Improve Existing Code

Best For

  • Code refactoring
  • Performance optimization
  • Adding features to existing code
  • Modernizing legacy code
Example usage: 
/enhance add loading states to forms
/enhance optimize database queries
/enhance add error handling
Expected output:
  • Enhanced code with improvements
  • Explanation of changes made
  • Before/after comparison
  • Tests for new functionality

​
/preview - Preview Changes Locally

Best For

  • Reviewing changes before committing
  • Testing features locally
  • Demonstrating work to stakeholders
Example usage: 
/preview start development server
/preview show recent changes

​
/status - Check Project Status

Best For

  • Understanding project health
  • Identifying issues or blockers
  • Getting overview of recent changes
Example usage: 
/status
/status show test coverage
/status check dependencies
Expected output:
  • Project health summary
  • Recent commits/changes
  • Test pass/fail status
  • Dependency status (outdated packages)
  • Known issues or warnings

​
/ui-ux-pro-max - Professional Design System

Best For

  • Creating polished, professional UIs
  • Need design inspiration
  • Want consistent design system
  • Client-facing applications
How it works:
  1. Loads ui-ux-pro-max skill (27 reference files)
  2. Provides access to 50 design styles
  3. Offers 21 color palettes
  4. Includes 50 font combinations
  5. Generates professional UI with selected aesthetic
Example usage: 
/ui-ux-pro-max create a landing page in minimalist style
/ui-ux-pro-max dashboard with glassmorphism design
Expected output:
  • Professional UI with cohesive design
  • Carefully selected color palette
  • Typography system with font pairings
  • Responsive layout
  • Accessibility-compliant components
  • Design tokens/variables
Browse the 50 available styles in the ui-ux-pro-max skill folder to see examples before requesting a specific aesthetic.

​
Chaining Commands

You can chain commands for complex workflows: 
1. /brainstorm authentication system
2. /plan implement JWT authentication
3. /orchestrate build authentication
4. /test authentication flow
5. /deploy to production
This workflow:
  1. Explores authentication options
  2. Creates implementation plan
  3. Builds the feature with multiple agents
  4. Tests the implementation
  5. Deploys to production

​
Command Selection Flowchart

Use this guide to choose the right command: 
Do you have clear requirements?
├─ No → /brainstorm
└─ Yes → Is it a new feature or existing code?
    ├─ New feature → Is it full-stack or single domain?
    │   ├─ Full-stack → /orchestrate
    │   └─ Single domain → /create
    ├─ Existing code → What do you want to do?
    │   ├─ Fix a bug → /debug
    │   ├─ Improve it → /enhance
    │   ├─ Test it → /test
    │   └─ Deploy it → /deploy
    └─ Large project → /plan first, then /orchestrate or /create

​
Pro Tips

Start with /brainstorm - Even if you think you know what you want, 5 minutes of brainstorming can save hours of rework.
Use /plan for estimates - Before committing to stakeholders, run /plan to get realistic time estimates.
Always /test before /deploy - Never skip testing, even for “small” changes. It’s much easier to fix bugs locally.
Combine /orchestrate with /ui-ux-pro-max - For professional full-stack apps, use orchestrate for structure and ui-ux-pro-max for design.

​
Next Steps

Build docs developers (and LLMs) love

Get started for freeTalk to us