ShopOps MCP

AI-powered server that implements the Model Context Protocol (MCP) for managing Shopify and WooCommerce stores.

Features

Store connectors for Shopify and WooCommerce.
12 MCP tools covering inventory, pricing, customers, orders, product performance and reporting.
4 MCP resources exposing store overview, inventory, recent orders and top customers.
Inventory forecasting using moving-average demand plus safety-stock calculation.
RFM-based customer segmentation (7 distinct segments).
Data-driven pricing analysis with margin-based optimization suggestions.
Order anomaly / fraud detection.
ABC analysis of product performance.
Automated daily and weekly reports.
Dual transport: local stdio and Streamable HTTP (MCPize).
TypeScript, @modelcontextprotocol/sdk v1.29+, Zod v4.
Free tier, plus a €29 lifetime Pro license.

Quick Start

bash

# 1. Install the package
npm i shopops-mcp

# 2. Create a .env file (see Configuration section)
cp .env.example .env

# 3. Run the server (local stdio mode)
npx shopops-mcp run --transport stdio

# 4. Or start the HTTP endpoint (MCPize deployment)
npx shopops-mcp run --transport http --port 8080

The server will read the environment variables, connect to the configured store(s), and expose the MCP tools and resources.

MCP Tools

Tool	Description
`store_connect`	Connects, lists, disconnects, or syncs a Shopify/WooCommerce store (`action: "connect" \| "list" \| "disconnect" \| "sync"`) and validates credentials.
`store_demo_seed`	Creates a realistic demo store (20 products, 40 customers, 150+ orders) so you can explore every tool without real store credentials.
`inventory_status`	Returns current stock levels, back-order flags and low-stock alerts.
`inventory_forecast`	Projects future inventory requirements using moving-average demand and safety-stock buffers.
`pricing_analyze`	Generates a price elasticity report and identifies under-/over-priced SKUs.
`pricing_optimize`	Suggests optimal price points based on margin analysis, sales velocity, and configurable pricing rules.
`customers_segment`	Performs RFM analysis and assigns customers to one of seven segments.
`customers_churn`	Scores customers for churn risk and provides retention recommendations.
`order_anomalies`	Detects potentially fraudulent or erroneous orders using pattern-recognition models.
`product_performance`	Conducts ABC analysis and returns contribution metrics per product class.
`report_daily`	Generates a JSON/CSV daily operations summary (sales, inventory, alerts).
`report_weekly`	Generates a weekly performance report with trend visualisations.

MCP Resources

Resource	Description
`store://overview`	High-level store metrics: product, order, and customer counts per connected store.
`store://inventory`	Low-stock alerts: active products with on-hand quantity ≤ 10, sorted lowest first.
`store://orders/recent`	The 20 most recent orders across all stores, with order number, total, status and date.
`store://customers/top`	Top 20 customers by total spending, with name, email and order count.

Configuration

ShopOps reads only a handful of environment variables. Store credentials are not env vars — they are passed to the store_connect tool at runtime (one connection per store), so the same server process can manage multiple Shopify/WooCommerce stores.

Variable	Required	Description
`PORT`	No	If set (or `MCPIZE=true`), the server runs as a Streamable HTTP endpoint on this port (default `8080`) instead of local `stdio`.
`MCPIZE`	No	Set to `true` by the MCPize runtime to force HTTP transport.
`LEMONSQUEEZY_LICENSE_KEY`	No	Pro license key. Without it the server runs in Free tier (see Pro License).

A minimal .env is provided in .env.example.

Connecting a store

Shopify and WooCommerce credentials are supplied as parameters to store_connect, e.g.:

jsonc

// Shopify
{ "action": "connect", "platform": "shopify",
  "store_domain": "myshop.myshopify.com", "access_token": "shpat_..." }

// WooCommerce
{ "action": "connect", "platform": "woocommerce",
  "store_url": "https://example.com", "consumer_key": "ck_...", "consumer_secret": "cs_..." }

store_connect returns a store_id that every other tool takes as input. To explore the server without real credentials, call store_demo_seed instead.

Roadmap (not yet implemented): report anonymization, S3 report export, and a configurable pricing-model / log-level are planned but not read by the current release.

Pro License

ShopOps ships in Free mode — store_demo_seed, store_connect, inventory_status, pricing_analyze, customers_segment, product_performance, and report_daily are open. The following tools require a Pro license:

inventory_forecast — moving-average demand forecasting + reorder points
pricing_optimize — actionable price-change recommendations
customers_churn — churn risk scoring + retention recommendations
order_anomalies — fraud / anomaly detection
report_weekly — week-over-week trend report + AI insights

Buy a Pro License (€29, lifetime, 3 machines): https://automatiabcn.lemonsqueezy.com/buy/cbbe44f0-a146-4c65-88c8-71f371037758

Or get the Indie MCP Stack Bundle (€69, all 4 servers).

server.ts

export LEMONSQUEEZY_LICENSE_KEY=YOUR-KEY-HERE

Or in your MCP client config:

config.json

{
  "mcpServers": {
    "shopops-mcp": {
      "command": "npx",
      "args": ["-y", "shopops-mcp-server"],
      "env": { "LEMONSQUEEZY_LICENSE_KEY": "YOUR-KEY-HERE" }
    }
  }
}

Validation is cached locally for 24 h — fully offline-capable after first run.

License

ShopOps MCP is released under the MIT License. See LICENSE for full terms.

Author: Automatia BCN