Documentation
Complete guide to AI Commands & Prompts on Commands.com
TL;DR - Quick Start
Commands.com lets you create, share, and monetize AI commands & prompts that work with Claude, ChatGPT, and other AI assistants.
- For Users: Browse and use AI commands/prompts with one-click install
 - For Creators: Use our MCP to generate commands.yaml automatically from your code
 - GitHub Integration: Import commands directly from your repositories
 - Monetization: Set prices and earn 85% revenue share via Stripe
 - MCP Support: Commands specify which MCP servers they need to function
 
Table of Contents
Commands.yaml Format
The commands.yaml file is the standard format for defining reusable AI commands and prompts with MCP server requirements. This format enables automatic synchronization with GitHub repositories and easy sharing across teams.
Multi-Command Structure
commands:
 # Example from ai-image-prompts repository
 - name: emergency-box
 title: ""In-Case-of-Emergency" Box Generator"
 description: "Generate emergency-style shadow box displays with customizable situations and objects"
 type: prompt
 category: "Image Generation"
 content: in-case-of-emergency-box/PROMPT.md
 license: LICENSE
 aiPlatform: "chatgpt-4o"
 readme: in-case-of-emergency-box/README.md
 inputParameters:
 - name: SITUATION
 label: "Emergency situation"
 description: "The type of emergency or crisis situation"
 type: "text"
 required: false
 default: "CREATIVE BLOCK"
 - name: MAIN_OBJECT
 label: "Main object in the box"
 description: "The primary item displayed in the emergency box"
 type: "text"
 required: false
 default: "vintage typewriter"
 # Example from ai-commands repository
 - name: code-reviewer
 title: AI Code Reviewer
 description: Intelligent code review assistant that analyzes code quality, security vulnerabilities, performance issues, and best practices
 type: command
 category: development
 commandName: "/code-review"
 aiPlatform: "claude-code@2025.06"
 content: code-reviewer/command.md
 readme: code-reviewer/README.md
 license: LICENSE
 thumbnail: "code-reviewer/preview.png"
 mcpRequirements:
 - serverId: github
 tier: required
 - serverId: filesystem
 tier: required
 - serverId: memory
 tier: required
 - serverId: sqlite
 tier: required
 inputParameters:
 - name: source
 label: Source Location
 description: GitHub URL or local file/directory path to analyze
 type: text
 required: true
 defaultValue: ""
 - name: focus
 label: Review Focus
 description: Primary area of analysis
 type: select
 required: false
 defaultValue: "all"
 options: ["security", "performance", "style", "all"]
# Global repository configuration
repository:
 license: LICENSE
 description: "A collection of AI commands and prompts"
 Required Fields
name
 Unique identifier for the command/prompt (used as title if no title specified)
 description
 Clear description of what this does
 type
 Either "command" for AI commands or "prompt" for AI prompts
 content
 Path to the main file (script or prompt template)
 Optional Fields
title
 Display title (overrides name for presentation, can include special characters)
 category
 Primary category (Image Generation, DevOps, Development, etc.)
 aiPlatform
 AI platform with version (claude-code@2025.06, chatgpt@4o, etc.)
 readme
 Path to detailed documentation
 license
 License file for this specific command
 inputs
 Input parameters for both commands and prompts with validation
 commandName
 For commands: The command name (e.g., /code-review, /project-setup)
 Unified Input Parameter System
Both AI commands and prompts use the same input parameter structure for consistency and better user experience:
inputParameters:
 # From ai-commands repository
 - name: source # Parameter identifier
 label: Source Location # Human-readable label
 description: GitHub URL or local file/directory path to analyze
 type: text # Input type (text, select)
 required: true # Whether parameter is required
 defaultValue: "" # Optional default value
 - name: focus
 label: Review Focus
 description: Primary area of analysis
 type: select
 required: false
 defaultValue: "all"
 options: ["security", "performance", "style", "all"]
 
 # From ai-image-prompts repository 
 - name: SITUATION
 label: "Emergency situation"
 description: "The type of emergency or crisis situation"
 type: "text"
 required: false
 default: "CREATIVE BLOCK"
 Supported Input Types
text Free-form text input ✅select Dropdown with predefined options (via GitHub import only) ⚠️path File or directory path (planned) 🚧number Numeric input with validation (planned) 🚧textarea Multi-line text input (planned) 🚧Current Limitations
Advanced input types are currently only supported via GitHub repository imports with commands.yaml files. The web form interface currently supports text inputs only. Full input type support in the web interface is planned.
MCP Server Requirements
Commands and prompts can specify required MCP (Model Context Protocol) servers that extend AI capabilities. This ensures users have the necessary tools before executing your command.
mcpRequirements:
 - serverId: "puppeteer" # Official MCP server ID
 tier: "required" # required | optional
 version: ">=1.0.0" # Optional version constraint
 - serverId: "sqlite"
 tier: "optional"
 description: "For data storage features"
 - serverId: "custom:analytics-server" # Custom server
 tier: "required"
 name: "Analytics MCP Server"
 description: "Custom analytics integration"
 customInstructions: "Install via: npm install @company/analytics-mcp"
 MCP Server Types
Official Servers From our curated directory (puppeteer, sqlite, github, etc.)Creator Servers Premium servers with subscription accessCustom Servers Use custom: prefix with installation instructionsVisual Assets & Gallery
Enhance your commands with thumbnails and gallery images that automatically sync from your repository. All images are processed through our optimization pipeline for fast loading across devices.
commands:
 # From ai-commands repository
 - name: code-reviewer
 title: AI Code Reviewer
 description: "Intelligent code review assistant that analyzes code quality..."
 content: code-reviewer/command.md
 thumbnail: "code-reviewer/preview.png" # Primary thumbnail
 # ... other fields
 
 # From ai-image-prompts repository
 - name: trading-card
 title: "Collectible Trading Card Generator"
 description: "Generate realistic trading cards in a shop display..."
 content: collectible-trading-card/PROMPT.md
 gallery: # Additional showcase images
 - "collectible-trading-card/images/image1.png"
 - "collectible-trading-card/images/image2.png"
 - "collectible-trading-card/images/image3.png"
 - "collectible-trading-card/images/image4.png"
 # ... other fields
 Image Specifications
Supported Formats
- • PNG: Best for screenshots, icons
 - • JPG/JPEG: Good for photos, complex images
 - • GIF: Animated demonstrations
 - • WebP: Modern format, smaller files
 
Automatic Processing
- • Thumbnails: 400x300, 800x600 generated
 - • Gallery: Multiple sizes for responsive display
 - • Optimization: Compressed for web delivery
 - • CDN: Global distribution via CloudFront
 
Complete YAML Examples
Here are complete examples from our actual repositories showing all available fields and features:
AI Commands Repository Example
name: AI Commands Collection
description: A curated collection of AI commands for development workflows and automation
version: 1.0.0
author: AI Commands Team
license: MIT
commands:
 - name: code-reviewer
 title: AI Code Reviewer
 description: Intelligent code review assistant that analyzes code quality, security vulnerabilities, performance issues, and best practices
 type: command
 category: development
 commandName: "/code-review"
 aiPlatform: "claude-code@2025.06"
 mcpRequirements:
 - serverId: github
 tier: required
 - serverId: filesystem
 tier: required
 - serverId: memory
 tier: required
 - serverId: sqlite
 tier: required
 content: code-reviewer/command.md
 readme: code-reviewer/README.md
 license: LICENSE
 thumbnail: "code-reviewer/preview.png"
 inputParameters:
 - name: source
 label: Source Location
 description: GitHub URL or local file/directory path to analyze
 type: text
 required: true
 defaultValue: ""
 - name: languages
 label: Programming Languages
 description: Specific languages to focus on (auto-detect if not specified)
 type: text
 required: false
 defaultValue: "auto"
 - name: focus
 label: Review Focus
 description: Primary area of analysis
 type: select
 required: false
 defaultValue: "all"
 options: ["security", "performance", "style", "all"]
 - name: save_report
 label: Save Report
 description: Generate detailed report files
 type: select
 required: false
 defaultValue: "yes"
 options: ["yes", "no"]
metadata:
 created: "2025-06-28"
 updated: "2025-06-28"
 total_commands: 3
 supported_platforms:
 - claude-code
 - claude
 minimum_requirements:
 - MCP support
 - File system access
 AI Image Prompts Repository Example
commands:
 - name: emergency-box
 title: ""In-Case-of-Emergency" Box Generator"
 description: "Generate emergency-style shadow box displays with customizable situations and objects"
 type: prompt
 category: "Image Generation"
 content: in-case-of-emergency-box/PROMPT.md
 license: LICENSE
 aiPlatform: "chatgpt-4o"
 readme: in-case-of-emergency-box/README.md
 inputParameters:
 - name: SITUATION
 label: "Emergency situation"
 description: "The type of emergency or crisis situation"
 type: "text"
 required: false
 default: "CREATIVE BLOCK"
 - name: MAIN_OBJECT
 label: "Main object in the box"
 description: "The primary item displayed in the emergency box"
 type: "text"
 required: false
 default: "vintage typewriter"
 - name: SUPPORT_ITEM_1
 label: "First supporting item"
 description: "Additional item to complement the main object"
 type: "text"
 required: false
 default: "stack of blank paper"
 - name: trading-card
 title: "Collectible Trading Card Generator"
 description: "Generate realistic trading cards in a shop display with custom characters and abilities"
 type: prompt
 category: "Image Generation"
 content: collectible-trading-card/PROMPT.md
 license: LICENSE
 aiPlatform: "chatgpt-4o"
 readme: collectible-trading-card/README.md
 gallery:
 - "collectible-trading-card/images/image1.png"
 - "collectible-trading-card/images/image2.png"
 - "collectible-trading-card/images/image3.png"
 - "collectible-trading-card/images/image4.png"
 inputParameters:
 - name: name
 label: "Card name"
 description: "The name printed on the trading card"
 type: "text"
 required: false
 default: "Accountant Squirrel"
 - name: character
 label: "Character description"
 description: "A brief description of the character shown on the card"
 type: "text"
 required: false
 default: "a very serious-looking grey squirrel"
# Global repository configuration
repository:
 license: LICENSE
 description: "A collection of creative AI image generation prompts"
 Live Examples
All examples above are from real working repositories. Check out ai-commands and ai-image-prompts for complete examples and to see the commands.yaml format in action.
GitHub Integration
Connect your GitHub repositories to automatically sync commands.yaml files. When you update files in your repo, they automatically appear on Commands.com.
Setup Process
- Go to your profile settings
 - Click "Connect GitHub Repository"
 - Authorize the Commands.com GitHub App
 - Select which repositories to sync
 - Add commands.yaml files to your repository
 
Repository Structure
my-repo/
├── commands/
│ ├── deploy-prod.yaml
│ ├── setup-env.yaml
│ └── backup-db.yaml
├── prompts/
│ ├── code-review.yaml
│ ├── documentation.yaml
│ └── bug-analysis.yaml
└── gallery/
 ├── deploy-screenshot.png
 └── setup-demo.gif
 Automatic Sync
- Changes to .yaml files trigger automatic updates
 - Gallery images are processed and optimized
 - Validation ensures proper format before publishing
 - Failed syncs send email notifications
 
Stripe Connect & Direct Payments
Direct Creator Payments
Creators receive 85% immediately, platform fee is automatically deducted
How Stripe Connect Works
- Connect your Stripe account through your creator profile
 - Complete identity verification (required by Stripe)
 - When users purchase your content, payment goes directly to you
 - Platform fee (15%) is automatically split during the transaction
 - You receive 85% of the sale price immediately
 
Payment Features
- Instant payouts: Money goes directly to your Stripe account
 - Global support: Available in 40+ countries
 - Multiple currencies: Accept payments in local currencies
 - Automatic tax handling: Stripe handles tax collection where required
 - Dispute protection: Stripe's fraud protection included
 - Real-time analytics: Track sales in your Stripe dashboard
 
Account Setup
Setting up Stripe Connect is straightforward and secure. Stripe handles all compliance and regulatory requirements, so you can focus on creating great content while they handle the payments infrastructure.
MCP Server Directory & Subscriptions
Commands.com hosts a comprehensive directory of Model Context Protocol (MCP) servers that extend AI capabilities. Our platform includes both curated free servers and creator-built premium servers with subscription access.
Curated Server Directory
All servers include tested installation commands and working documentation
Popular Categories
- Web Automation: Puppeteer, Playwright
 - File Systems: Filesystem, SQLite
 - Memory & Data: Memory, PostgreSQL
 - Development: GitHub, Git
 - Communication: Discord, Slack
 - APIs: Fetch, Sequential Thinking
 
Installation Process
- Browse the MCP server directory
 - Copy the installation command
 - Run in your terminal/environment
 - Configure with Claude Code or your AI assistant
 - Start using enhanced AI capabilities
 
Creator MCP Server Subscriptions
Creators can publish specialized MCP servers with subscription-based access, enabling monetization of custom integrations and advanced functionality.
For Creators
- • Publish custom MCP servers
 - • Set subscription pricing tiers
 - • Earn 85% revenue share
 - • Manage subscribers
 
For Subscribers
- • Access premium MCP servers
 - • Specialized integrations
 - • Enhanced AI capabilities
 - • Cancel anytime
 
Popular Use Cases
- • Custom API integrations
 - • Enterprise databases
 - • Workflow automation
 - • Specialized tools
 
Creating Great Content
AI Commands Best Practices
- Use clear command naming conventions (/security-audit, /code-review)
 - Define input parameters with descriptions and defaults
 - Specify compatible AI platform versions
 - Provide usage examples and expected outputs
 - Include context and constraints for AI execution
 - Test with target AI platform before publishing
 
AI Prompts Best Practices
- Be specific about the desired output format
 - Include context and constraints
 - Test with multiple AI models
 - Provide example inputs and outputs
 - Specify which AI platforms work best
 
Gallery Images
- Show the command/prompt in action
 - Include before/after screenshots
 - Use GIFs for dynamic demonstrations
 - Optimize file sizes (max 5MB)
 - Support: PNG, JPG, GIF, WebP
 
Platform Features
Search & Discovery
- Full-text search across all content
 - Category and tag filtering
 - Trending and popular sections
 - Creator profiles and portfolios
 
User Features
- Like and copy tracking
 - Purchase history
 - Personal dashboard
 - Content recommendations
 
Creator Tools
- Analytics and insights
 - Revenue tracking
 - Content management
 - GitHub synchronization
 
Quality Control
- Automated content validation
 - Community moderation
 - Security scanning
 - Spam prevention
 
Best Practices
✓ Do
- • Write clear, descriptive titles
 - • Include comprehensive documentation
 - • Test thoroughly before publishing
 - • Use appropriate categories and tags
 - • Provide visual examples
 - • Update content regularly
 - • Respond to user feedback
 
✗ Don't
- • Submit untested or broken code
 - • Use misleading descriptions
 - • Include malicious content
 - • Violate platform terms
 - • Spam tags or categories
 - • Copy others' work without credit
 - • Ignore security best practices
 
Need Help?
Can't find what you're looking for? Our community is here to help.
Join Community Support