Python · Model Context Protocol · Shopify Admin GraphQL
Shopify MCP — production-ready product data for AI
I built an open-source, read-only MCP connector that gives AI assistants secure access to Shopify product, variant, inventory and order data. It works across product categories through Shopify’s standard product model and supports local use or a hosted, multi-tenant deployment with merchant OAuth, encrypted credentials and bearer-protected MCP tools.
How I use it with inriver: inriver remains the source of enriched product information and syndicates each approved update to Shopify. The MCP connector then reads the channel-ready Shopify record and anchors AI responses to the latest published product state—supporting product verification, variant and inventory summaries, merchandising questions and downstream AI workflows.
Production capabilities: OAuth and token rotation, encrypted tenant storage, cursor pagination, retry/backoff, uninstall cleanup, structured logs, metrics, health probes and Docker deployment. Large multi-replica or regulated deployments still require the documented infrastructure and compliance controls.
Read the engineering case study View GitHub
