Cursor

Why Cursor + Synchronity?

Cursor's agent mode can call MCP tools. Wire Synchronity in and your coding agent can:

• Look up real product data while writing storefront integration code
• Place test orders against a sandbox site to verify your checkout flow end-to-end
• Inspect order history when debugging fulfilment bugs

Setup

Cursor reads MCP servers from ~/.cursor/mcp.json (or per-project .cursor/mcp.json).

{
  "mcpServers": {
    "synchronity": {
      "command": "npx",
      "args": ["-y", "@synchronity/mcp-server"],
      "env": {
        "GATEWAY_URL": "https://api.synchronity.app",
        "AIT_TOKEN": "<your AIT>",
        "DEFAULT_SITE_ID": "<sandbox site id>"
      }
    }
  }
}

Reload Cursor (Cmd+Shift+P → "Reload Window"). Synchronity tools appear in the agent's tool palette.

Recommended setup for development

Use a sandbox site (a separate Shopify dev store or a local WooCommerce instance) when running Synchronity through a coding agent. Wire its site_id into DEFAULT_SITE_ID so the agent doesn't accidentally hit production.

Use a dedicated AIT with only read_products + manage_cart scopes during normal coding sessions. Issue a separate token with execute_checkout only when you're explicitly testing checkout flows.

Delegation in Cursor

Cursor doesn't render rich verification UIs the way Claude does — the verification URL surfaces as plain text in the agent output. Open it in your browser, approve, return to Cursor. The agent polls and continues.