Savvy Scratch MCP Server

Install

Claude Desktop

Open your claude_desktop_config.json:

• macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
• Windows: %APPDATA%\Claude\claude_desktop_config.json

Add the Savvy Scratch server:

{
  "mcpServers": {
    "savvyscratch": {
      "command": "npx",
      "args": ["-y", "@savvyscratch/mcp-server"]
    }
  }
}

Restart Claude Desktop. You're done.

Claude Code

Add to your project's .mcp.json (or run claude mcp add):

{
  "mcpServers": {
    "savvyscratch": {
      "command": "npx",
      "args": ["-y", "@savvyscratch/mcp-server"]
    }
  }
}

Other MCP-compatible clients

Any client that supports stdio MCP servers can run:

npx -y @savvyscratch/mcp-server

What is this?

An MCP (Model Context Protocol) server that gives AI assistants access to Savvy Scratch — pro-gambler-grade analysis of scratch-off lottery games across 19 U.S. states.

Register, sign in, browse live game data, and subscribe — all through natural conversation with your AI. Zero configuration required.

First-time setup (through your AI)

Once installed, just talk to your AI:

1. "Register a Savvy Scratch account for me" → AI uses the register tool
2. "What scratch-off games should I play in Florida?" → AI uses login + get_games
3. "Subscribe me to Savvy Scratch" → AI uses get_subscribe_link

No API keys. No env vars. Your account lives in your AI session.

Tools

Auth & Account

Game Analysis

Pricing

Free accounts see top-3 previews in every state. A Savvy Scratch subscription ($5/mo or $50/yr) unlocks:

• All games in all 19 states
• Detailed prize-tier breakdowns with live remaining prizes
• Budget-optimized recommendations ranked by weighted score
• Pro gambler methodology applied to every game

Example Conversation

You: What scratch-off games should I play in Florida?

AI: [calls login, then get_games with state=fl]
    Shows top 3 games with scores and odds...
    "Subscribe to see all 100 games. Use get_subscribe_link to get started."

You: I have $10 to spend in Georgia, what's best?

AI: [calls best_games with state=ga, budget=10]
    Shows top-rated games at $10 or under ranked by weighted score

You: Show me the prize breakdown for Atlanta Falcons

AI: [calls game_details with state=ga, game_name="Atlanta Falcons"]
    Full prize tier breakdown with remaining prizes and adjusted odds

Build from source

git clone https://github.com/myque08/savvymcp.git
cd savvymcp
npm install
npm run build
node dist/index.js

Links

• npm package: https://www.npmjs.com/package/@savvyscratch/mcp-server
• Website: https://www.savvyscratch.com
• GitHub: https://github.com/myque08/savvymcp
• Report issues: https://github.com/myque08/savvymcp/issues