Pilates With Neelam MCP Server

An MCP (Model Context Protocol) server that connects AI agents (ChatGPT, Claude, Cursor, etc.) to the Pilates With Neelam studio management platform. Provides real-time access to classes, bookings, student profiles, assessments, and more.

Quick Start

# Install dependencies
npm install

# Build
npm run build

# Run (stdio mode for MCP hosts)
npm start

Related MCP server: @aiwerk/mcp-server-cal

Studio Identity

This MCP server is configured exclusively for Pilates With Neelam (studio_id: pilates_with_neelam). All tools verify the studio identity and reject requests for any other studio.

Tools

Configuration for ChatGPT Desktop

1. Build the server: npm run build

2. Add to your ChatGPT MCP config (~/Library/Application Support/ChatGPT/mcp_servers.json on macOS, %APPDATA%\ChatGPT\mcp_servers.json on Windows):

{
  "mcpServers": {
    "pilates-with-neelam": {
      "command": "node",
      "args": ["C:\\Users\\Welcome\\pilates-with-neelam-mcp\\dist\\index.js"],
      "env": {
        "SUPABASE_URL": "https://ywyagyftnwnybrxtyhjd.supabase.co",
        "SUPABASE_ANON_KEY": "sb_publishable_...",
        "SUPABASE_SERVICE_ROLE_KEY": "sb_secret_...",
        "STUDIO_ID": "pilates_with_neelam",
        "STUDIO_NAME": "Pilates With Neelam",
        "STUDIO_DOMAIN": "pilateswithneelam.in"
      }
    }
  }
}

Configuration for Claude Desktop

Add to ~/Library/Application Support/Claude/claude_desktop_config.json (macOS) or %APPDATA%\Claude\claude_desktop_config.json (Windows):

{
  "mcpServers": {
    "pilates-with-neelam": {
      "command": "node",
      "args": ["C:\\Users\\Welcome\\pilates-with-neelam-mcp\\dist\\index.js"],
      "env": {
        "SUPABASE_URL": "https://ywyagyftnwnybrxtyhjd.supabase.co",
        "SUPABASE_ANON_KEY": "sb_publishable_...",
        "SUPABASE_SERVICE_ROLE_KEY": "sb_secret_...",
        "STUDIO_ID": "pilates_with_neelam",
        "STUDIO_NAME": "Pilates With Neelam",
        "STUDIO_DOMAIN": "pilateswithneelam.in"
      }
    }
  }
}

Configuration for Cursor

Add to .cursor/mcp.json in your project:

{
  "mcpServers": {
    "pilates-with-neelam": {
      "command": "node",
      "args": ["C:\\Users\\Welcome\\pilates-with-neelam-mcp\\dist\\index.js"],
      "env": {
        "SUPABASE_URL": "https://ywyagyftnwnybrxtyhjd.supabase.co",
        "SUPABASE_ANON_KEY": "sb_publishable_...",
        "SUPABASE_SERVICE_ROLE_KEY": "sb_secret_...",
        "STUDIO_ID": "pilates_with_neelam",
        "STUDIO_NAME": "Pilates With Neelam",
        "STUDIO_DOMAIN": "pilateswithneelam.in"
      }
    }
  }
}

Complete Workflow Demo

An AI agent can perform this real workflow using live data:

1. "What studio is this?" → get_studio_manifest → Returns studio identity

2. "Find me a reformer class" → find_my_class → Returns matching classes with availability

3. "What's the assessment for a beginner?" → assess_pilates_level → Returns level + recommendations

4. "Check availability for class c_m1abc" → get_pilates_with_neelam_availability → Returns real-time spots

5. "Book student s2 into that class" → book_pilates_with_neelam_class → Returns booking confirmation with studio ID

6. "Log check-in for s2: felt great, core strong" → check_in_pilates_with_neelam_class → Returns AI summary + exercises

Environment Variables

Testing

# Run end-to-end tests
npm run test:e2e

Architecture

• Transport: stdio (MCP standard)

• Database: Supabase (Postgres + RLS)

• Auth: Service role key for server-to-server, optional auth token for user-scoped operations

• Studio Isolation: All tools verify studio_id before operations

• Privacy: No student PII (names, emails, phones) is exposed. Tools return only class and availability data.

License

MIT