Skip to content
Reponse.ai LogoDocs
DocsAPISDKsGuides

Search Documentation

Search for an article...

Getting Started

  • Overview
  • Architecture
  • Installation
  • Authentication
  • First Request

Products & Collections

  • Products — List
  • Products — Get
  • Products — Metafields
  • Collections — List
  • Collections — Get
  • Collections — Products

Cart & Checkout

  • Cart — Create
  • Cart — Get
  • Cart — Add Item
  • Cart — Update Item
  • Cart — Remove Item
  • Cart — Apply Promotion
  • Checkout — Stripe
  • Checkout — Payment Intent
  • Checkout — ACP

Orders & Fulfillment

  • Orders — Create
  • Orders — Get
  • Orders — Confirm
  • Orders — Cancel
  • Orders — Fulfill
  • Orders — Refund

Commerce

  • Inventory — Get
  • Inventory — Update
  • Shipping — Rates
  • Discounts — List
  • Discounts — Create
  • Discounts — Validate
  • Subscriptions — Manage

Loyalty & Gift Cards

  • Loyalty — Balance
  • Loyalty — Redeem
  • Loyalty — Referral
  • Gift Cards — List
  • Gift Cards — Redeem

Tickets & Support

  • Tickets — List
  • Tickets — Create
  • Tickets — Reply

Platform

  • Product Feed — JSON
  • Product Feed — CSV
  • Theme — Get
  • Approvals — Execute
  • Approvals — Reject
  • Geocode

SDKs

  • SDK Overview
  • TypeScript SDK
  • React Hooks

Guides

  • Chat Widget
  • Shopify Sync
  • Storefront Starter
  • Loyalty Program
  • Discounts & Promotions
  • Subscriptions
  • Klaviyo Integration
  • Custom Domains
  • Agentic Commerce (ACP)
  • A2A Protocol
  • AI Engines
  • MCP Server

Webhooks

  • Webhooks Overview
  • Events Reference
  • Shopify Webhooks
  • Stripe Webhooks
  • Reviews (Stamped / Trustpilot)
  • Logistics
  • Email Inbound

Resources

  • Environment Variables
  • Rate Limits
  • Changelog
DocsGuidesMCP Server

MCP Server

Connect AI tools to Reponse via the MCP server.

2 min read/Last updated Jul 9, 2026
On this page

Overview

Reponse ships a **Model Context Protocol (MCP)** server that exposes commerce operations as typed tools to any MCP-compatible client — including Claude, Cursor, Windsurf, and custom agent orchestrators. This lets an AI agent browse the catalog, manage carts, check orders, and act on loyalty data without building a custom integration.

Prerequisites

RequirementDescription
Reponse workspaceActive workspace with an API key
MCP-compatible clientClaude Desktop, Cursor, or any MCP SDK client
API key scopecatalog:read, cart:write, orders:read (minimum)

Step 1 — Configure the MCP client

Add the Reponse MCP server to your client configuration. For **Claude Desktop**, edit `claude_desktop_config.json`:

json
{
  "mcpServers": {
    "reponse": {
      "command": "npx",
      "args": ["-y", "@reponseai/mcp"],
      "env": {
        "REPONSE_API_KEY": "rp_live_xxxxxxxxxxxx"
      }
    }
  }
}

For **Cursor**, add the same entry under `.cursor/mcp.json`:

json
{
  "mcpServers": {
    "reponse": {
      "command": "npx",
      "args": ["-y", "@reponseai/mcp"],
      "env": {
        "REPONSE_API_KEY": "rp_live_xxxxxxxxxxxx"
      }
    }
  }
}

Step 2 — Verify the connection

Once configured, your MCP client discovers the available tools automatically. You should see tools prefixed with `reponse_` in your client's tool list.

Available tools

The MCP server exposes the following tools, mapping directly to the `/v1` REST API:

ToolDescriptionAPI equivalent
reponse_list_productsSearch and list products with filtersGET /v1/catalog/products
reponse_get_productGet full product details by IDGET /v1/catalog/products/:id
reponse_list_collectionsList all collectionsGET /v1/catalog/collections
reponse_create_cartCreate a new cartPOST /v1/cart
reponse_add_to_cartAdd an item to a cartPOST /v1/cart/:id/items
reponse_get_cartRetrieve cart contentsGET /v1/cart/:id
reponse_apply_promoApply a promo code to a cartPOST /v1/cart/:id/promo
reponse_list_ordersList orders by email or contactGET /v1/orders
reponse_get_orderGet order detailsGET /v1/orders/:id
reponse_get_loyalty_balanceCheck loyalty points balanceGET /v1/loyalty/:contactId/balance
reponse_create_ticketOpen a support ticketPOST /v1/tickets

Tool input/output schema

Each tool uses typed JSON Schema. Example for `reponse_list_products`:

json
{
  "name": "reponse_list_products",
  "description": "Search and list products from the catalog",
  "inputSchema": {
    "type": "object",
    "properties": {
      "query":  { "type": "string", "description": "Search query" },
      "limit":  { "type": "number", "default": 10 },
      "cursor": { "type": "string" }
    }
  }
}

Step 3 — Use in an agent workflow

With the MCP server connected, an agent can chain tools naturally:

User: "Find me a red jacket under €100 and add it to my cart"

Agent:
  1. reponse_list_products({ query: "red jacket", limit: 5 })
  2. reponse_create_cart({ currency: "EUR" })
  3. reponse_add_to_cart({ cartId: "...", variantId: "...", quantity: 1 })

Configuration reference

VariableRequiredDescription
REPONSE_API_KEYYesWorkspace API key
REPONSE_BASE_URLNoOverride API base URL (default: https://api.reponse.ai)
MCP_LOG_LEVELNoLogging verbosity: debug, info, warn, error

A2A integration

The MCP server can be used alongside the [A2A protocol](/docs/a2a-protocol) to let external agents interact with Reponse. An orchestrator agent can discover the Reponse agent card via A2A, then invoke tools through MCP for catalog and cart operations.

Troubleshooting

SymptomCauseFix
Tools not appearing in clientConfig file path wrongVerify config location for your MCP client
UNAUTHORIZED errorsInvalid API keyCheck REPONSE_API_KEY env variable
npx command failsNode.js not installed or outdatedInstall Node.js 18+
Slow tool responsesNetwork latency or large catalogReduce limit parameter in list calls
Tool returns empty resultsNo products in workspaceAdd products via Shopify sync or dashboard
PreviousAI Engines
NextWebhooks Overview