Getting Started with UCM
Get up and running with UCM in under 1 minute. No downloads, no setup files, no API keys to manage.
3 Simple Steps
Create your free account
Quick sign up
Verify your email
Check your inbox and click the link
Connect to Claude Desktop
Add Custom Connector in settings
Connection Details:
(Authentication happens automatically)
Method 1:
File → Settings → Connectors
Method 2:
Click "Search and Tools" → "Add Connector"
That's it! No downloads, no setup files, no API keys to manage.
For Developers
Advanced Setup: Local MCP Server
The Local MCP Server brings performance enhancements, especially when working with large files.
What You'll Need
- • Node.js 18+ installed on your system
- • An AI assistant that supports MCP
- • About 5 minutes for setup
Advanced Setup Process
For developers who need enhanced performance with large files.
Account Setup
Create your free UCM account
Sign Up for a Free Account
Create your free UCM account to store and manage your guidance artifacts:
Sign Up for UCM →You'll receive an email verification link to complete account setup.
Configure your MCP Server
Connect your AI assistant to UCM
Get Your Local MCP Configuration
Generate your auth token and get ready-to-use local MCP server configurations:
Get Local MCP Config →What You'll Get:
- • Your unique Author ID and auth token
- • Copy-paste ready Claude Desktop configuration
- • Claude Code CLI setup command
- • Complete local MCP server connection details
Install Local MCP Server
The local UCM MCP Server uses Node.js and will download the ucm-mcp-server package. Update your AI assistant configuration using the details from the API Keys page.
Supported AI Assistants:
- • Claude Desktop - JSON configuration for MCP servers
- • Claude Code CLI - Command line setup
- • Continue.dev - VS Code extension
- • Any MCP-compatible AI client
Important: Restart your AI assistant after adding the MCP configuration to ensure the server connection loads properly.
Publish and Discover
Instruct your AI
Mandatory Starting Prompt
Use this prompt for your AI to easily use the MCP Server:
Expected Response
Your AI assistant should respond with UCM quickstart information, confirming the MCP server is connected and working properly. Your AI should also read your Author Index, which gives it key details about working with your own Context.
🎉 You're Ready!
Your UCM environment is configured. Now you can now store your own guidance and artifacts, or build applications with public Utaba AI guidance, proven patterns, and re-usable code.
🚀 Build Your First App
Follow our step-by-step guide to create a new application using UCM guidance and micro-block patterns.
Start Building →📚 Browse UCM Content
Explore the public utaba repository with battle-tested patterns, guidance, and implementations.
Browse Content →💡 Remember
Always start your AI conversations with "Read the UCM quickstart" to ensure your AI assistant understands the UCM context and follows proper workflows.