MCP Server for AI Assistants

Transform AI assistants like Claude Desktop into expert e-commerce strategists with instant access to your store data.

What is MCP?

Model Context Protocol (MCP) is Anthropic's official standard for connecting AI assistants to external data sources and tools. The AIVA MCP Server gives Claude (and other AI assistants) the ability to:

✅ Query your customer data in natural language ✅ Analyze RFM segments and trends ✅ Calculate MRR, LTV, churn rates ✅ Create prize draws conversationally ✅ Generate marketing campaign strategies ✅ Provide ROI-based recommendations

Quick Setup with Claude Desktop

1. Install the MCP Server


# Option A: Use NPM package (when published)

npm install -g @getaiva/mcp

# Option B: Clone from source

git clone https://github.com/growth-spurt/aiva-mcp-server.git
cd aiva-mcp-server
npm install
npm run build

2. Configure Claude Desktop

Mac: Edit ~/Library/Application Support/Claude/claude_desktop_config.json Windows: Edit %APPDATA%\Claude\claude_desktop_config.json

{
  "mcpServers": {
    "aiva": {
      "command": "npx",
      "args": ["-y", "@getaiva/mcp"],
      "env": {
        "AIVA_API_KEY": "your-api-key-here",
        "AIVA_MERCHANT_ID": "your-merchant-id-here",
        "AIVA_API_BASE_URL": "https://api.getaiva.ai"
      }
    }
  }
}

3. Restart Claude Desktop

Quit and reopen Claude Desktop for changes to take effect.

4. Start Asking Questions!

You: "Show me my top 20 customers by lifetime value"

Claude: [Queries AIVA API]
        Here are your top 20 customers:

        1. John Smith - $2,450.75 (Champions segment, 15 orders)
        2. Sarah Johnson - $2,234.20 (Champions segment, 12 orders)
        ...

        Key insights:
        • Top 20 customers represent 35% of total revenue
        • Average LTV: $1,847
        • All are in Champions or Loyal segments

        Recommendation: Create a VIP loyalty program for this group.

You: "How many at-risk customers do I have?"

Claude: You have 87 at-risk customers representing $145,320 in lifetime value.

        This is concerning - they used to be good customers but haven't
        ordered in 30-60 days. I recommend:

        1. Win-back email campaign with 20% discount
        2. Expected recovery rate: 15-25% (13-22 customers)
        3. Potential recovered revenue: $21,798 - $36,330
        4. Campaign cost estimate: $500
        5. ROI: 43x to 72x return

        Want me to help create the email campaign content?

You: "Calculate my MRR and ARR"

Claude: Your current subscription metrics:

        📊 MRR (Monthly Recurring Revenue): $17,095.50
        📊 ARR (Annual Recurring Revenue): $205,146
        📊 Active Subscriptions: 342
        📊 ARPU (Avg Revenue Per User): $49.99

        Growth trends:
        • MRR up 12% vs last month
        • 28 new subscriptions this month
        • Churn rate: 3.2% (industry avg: 5-7%)

        You're performing well above industry benchmarks!

Available MCP Resources

Resources that Claude can query:

Resource	Data Returned
`aiva://merchant/info`	Store profile, settings, subscription tier
`aiva://customers/list`	All customers with RFM segments
`aiva://subscriptions/list`	Active subscriptions and billing
`aiva://affiliates/list`	Affiliate performance metrics
`aiva://prize-draws/list`	Prize draw campaigns
`aiva://orders/list`	Order history and sales data
`aiva://intelligence/overview`	Complete analytics dashboard

Available MCP Tools

Actions that Claude can perform:

Tool	What It Does
`get_customer`	Get detailed customer info
`search_customers`	Search by RFM segment
`get_subscription`	Get subscription details
`sync_subscriptions`	Trigger sync from Shopify
`get_affiliate`	Get affiliate commission history
`create_prize_draw`	Create blockchain prize draw
`get_orders`	Get orders with date filtering
`send_aiva_message`	Send AI message to customer

Complete Documentation