← Architecture Overview | Next: Database Schema and Data →

Step 1: Install Prerequisites

Install the Supabase CLI:

# macOS (Homebrew)
brew install supabase/tap/supabase
 
# npm (cross-platform)
npm install -g supabase
 
# Verify installation
supabase --version

Ensure you have:

  • Docker Desktop running (required for local development)
  • Node.js 20+ installed
  • A Supabase account (free tier is fine)

Step 2: Create a New Supabase Project

Option A: Cloud Project (recommended for deploying)

  1. Go to supabase.com/dashboard
  2. Click "New Project"
  3. Choose an organization (or create one)
  4. Set a project name: mcp-inventory
  5. Set a database password (save this somewhere safe)
  6. Choose a region close to you
  7. Click "Create new project"

Once created, note your:

  • Project Reference ID (from the URL: https://supabase.com/dashboard/project/<project-ref>)
  • Project URL: https://<project-ref>.supabase.co
  • Service Role Key: Found in Settings > API (keep this secret!)

Option B: Local-Only Development

You can develop entirely locally without a cloud project:

mkdir mcp-inventory
cd mcp-inventory
supabase init
supabase start

This spins up a full local Supabase stack (Postgres, Auth, Storage, Edge Functions) via Docker.

Step 3: Initialize the Project Structure

mkdir mcp-inventory
cd mcp-inventory
supabase init

This creates:

mcp-inventory/
├── supabase/
│   ├── config.toml          # Supabase project configuration
│   ├── functions/            # Edge Functions live here
│   ├── migrations/           # Database migrations
│   └── seed.sql              # Seed data (we'll create this)

Step 4: Create the Edge Function

supabase functions new mcp

This creates supabase/functions/mcp/index.ts with a starter template. We will replace this with our MCP server code shortly.

Your project structure should now look like:

mcp-inventory/
├── supabase/
│   ├── config.toml
│   ├── functions/
│   │   └── mcp/
│   │       └── index.ts      # Our MCP server (we'll write this)
│   ├── migrations/            # SQL migrations (we'll create these)
│   └── seed.sql               # Dummy data (we'll create this)

Next: Module 4 - Creating the Database Schema and Dummy Data →