AI Tools for Varity

Varity Team Core Contributors Updated May 2026

Varity is built to be driven from your AI coding tool. Add the MCP server once, then ask your AI assistant to deploy, monitor, and manage your apps in plain English.

Available Tools

MCP Server

Model Context Protocol server for Cursor, Claude Code, VS Code, and Windsurf. Deploy, search docs, and manage apps from your AI editor.

View Spec →

LLM-Optimized Docs

Machine-readable documentation for AI assistants. Available in two formats: /llms.txt (summary) and /llms-full.txt (detailed).

Quick Start: MCP Server

The Varity MCP server lets your AI editor create, deploy, and manage Varity apps directly.

For Cursor: add to .cursor/mcp.json:

{
  "mcpServers": {
    "varity": {
      "command": "npx",
      "args": ["-y", "@varity-labs/mcp@beta"]
    }
  }
}

For Claude Code:

claude mcp add varity -- npx -y @varity-labs/mcp@beta

Then ask your AI to “deploy this project” or “migrate this app from Vercel.” The MCP server provides a full set of tools for deploying, monitoring, and managing your apps. See the MCP Server Spec for the complete tool reference, or check out the Build with AI (No Code) tutorial to get started.

LLM-Optimized Documentation

AI assistants can reference our documentation in machine-readable formats:

File	URL	Use Case
`llms.txt`	`https://docs.varity.so/llms.txt`	Summary: what Varity is, key APIs, common patterns
`llms-full.txt`	`https://docs.varity.so/llms-full.txt`	Detailed: full documentation content for deep context

Point any AI tool at one of these URLs for instant Varity context. Use llms.txt for quick context in chat-based LLMs (ChatGPT, Gemini), and llms-full.txt when the AI needs comprehensive documentation.

AI-Friendly Patterns

When you point an AI tool at Varity’s LLM-optimized docs, it learns to use developer-friendly language:

“Deploy your application” (not platform-specific hosting terms)
“Sign in” (the user just sees a normal login screen)
“Your backend is configured automatically” (no infrastructure jargon)

It also learns the real deploy flow: detect the framework, choose the hosting mode, auto-wire any databases or caches your app needs, and return a live URL, all without changing your code.

Supported AI Tools

Tool	Integration	Level
Cursor	MCP Server + `llms.txt` context	Full
Claude Code	MCP Server + `llms.txt` context	Full
VS Code (Copilot)	MCP Server + `llms.txt` context	Full
Windsurf	MCP Server + `llms.txt` context	Full
ChatGPT	`llms.txt` context	Basic
Gemini	`llms.txt` context	Basic
GitHub Copilot	`llms.txt` context	Basic

What’s Next

MCP Server Spec: Setup guide and tool reference
Build with AI (No Code): Deploy your first app from your AI editor