schneider-ops-mcp

Cross-system orchestration MCP server for Schneider Saddlery — connects Shopify Plus and Fulfil.io ERP into a unified AI-accessible operations layer.

What This Does

This MCP server gives AI agents real-time access to both the storefront (Shopify) and the warehouse/ERP (Fulfil.io) simultaneously. Instead of switching between dashboards, an AI assistant can:

• Reconcile inventory discrepancies between systems
• Trace orders across the full pipeline (Shopify → Fulfil → Shipment → Tracking)
• Generate a unified morning ops dashboard with cross-system health scoring
• Forecast demand and generate proactive restock alerts
• Monitor shipping SLAs with carrier performance breakdowns

Tools (14)

Inventory Reconciliation

| Tool | Description | |------|-------------| | inventory_reconciliation | Compare inventory for a single SKU across both systems | | stock_discrepancy_report | Scan all products for inventory mismatches | | sync_inventory_to_shopify | Calculate adjustments to align Shopify with Fulfil (dry run) |

Order Pipeline

| Tool | Description | |------|-------------| | order_pipeline_status | Trace a single order across both systems end-to-end | | order_exception_monitor | Find orders with cross-system issues | | delayed_fulfillment_report | Find stuck orders and identify bottlenecks |

Unified Views

| Tool | Description | |------|-------------| | unified_product_view | Full product view combining storefront + ERP data | | unified_customer_view | Customer profile combining Shopify + Fulfil data | | unified_daily_dashboard | Comprehensive morning operations dashboard |

Analytics

| Tool | Description | |------|-------------| | revenue_reconciliation | Compare revenue between systems for a date range | | catalog_sync_check | Verify product catalog consistency | | seasonal_demand_forecast | Project demand based on historical order velocity |

Automation

| Tool | Description | |------|-------------| | generate_restock_alerts | Identify products nearing stockout with restock recommendations | | shipping_sla_monitor | Track order-to-ship time and carrier performance |

Setup

Prerequisites

• Node.js 18+
• Fulfil.io API key with read access
• Shopify Admin API access token

Environment Variables

FULFIL_API_KEY=your-fulfil-api-key
FULFIL_SUBDOMAIN=schneider-saddlery    # → https://schneider-saddlery.fulfil.io
SHOPIFY_STORE_DOMAIN=schneider-saddlery.myshopify.com
SHOPIFY_ACCESS_TOKEN=shpat_xxxxx

Install & Build

npm install
npm run build

Run

# Direct
node dist/index.js

# Via wrapper (loads .env automatically)
./schneider-ops-mcp-wrapper.sh

MCP Configuration

Add to your Claude Code settings:

{
  "mcpServers": {
    "schneider-ops": {
      "command": "/Volumes/Virtual Server/projects/schneider-ops-mcp/schneider-ops-mcp-wrapper.sh",
      "env": {
        "FULFIL_API_KEY": "your-key",
        "FULFIL_SUBDOMAIN": "schneider-saddlery",
        "SHOPIFY_STORE_DOMAIN": "schneider-saddlery.myshopify.com",
        "SHOPIFY_ACCESS_TOKEN": "shpat_xxxxx"
      }
    }
  }
}

Architecture

schneider-ops-mcp
├── Shopify Admin REST API (2024-10) ←→ ShopifyClient
│   ├── Orders, Products, Inventory, Customers
│   ├── Rate limiting via X-Shopify-Shop-Api-Call-Limit
│   └── Pagination via Link headers
│
├── Fulfil.io REST API v2 ←→ FulfilClient
│   ├── Sales, Products, Shipments, Parties
│   ├── Tryton-style domain filters
│   └── Bearer token auth
│
└── 14 Cross-System Tools
    ├── Match products by SKU
    ├── Match customers by email
    └── Match orders by reference/order number

Both API clients include exponential backoff retries, rate limit handling, and structured error reporting.

License

MIT