Skip to main content
Glama

Etsy MCP Server for OpenAI Agent Builder

This is a deployable remote MCP server that lets OpenAI Agent Builder connect directly to Etsy using Etsy OAuth 2.0 with PKCE.

Tools

  • etsy_get_my_shops

  • etsy_get_shop

  • etsy_get_listings

  • etsy_create_listing

  • etsy_update_listing

  • etsy_get_orders

  • etsy_get_messages

  • etsy_reply_to_message

Important: Etsy's official OpenAPI v3 spec does not expose conversations/messages or reply endpoints. The two message tools are registered so Agent Builder can see the requested surface, but they return unsupported_by_etsy_open_api_v3 instead of faking unsupported Etsy calls.

Related MCP server: MCP OAuth Server

Environment Variables

PUBLIC_BASE_URL=https://your-deployed-etsy-mcp.example.com
PORT=3000
DATA_PATH=etsy-mcp-store.json
ETSY_CLIENT_ID=your_etsy_keystring
ETSY_API_KEYSTRING=your_etsy_keystring
ETSY_SHARED_SECRET=your_etsy_shared_secret

ETSY_CLIENT_ID and ETSY_API_KEYSTRING are normally the same Etsy keystring.

Etsy Developer Portal Setup

  1. Go to https://www.etsy.com/developers/your-apps.

  2. Create or open your Etsy app.

  3. Copy the app keystring into ETSY_CLIENT_ID and ETSY_API_KEYSTRING.

  4. Copy the shared secret into ETSY_SHARED_SECRET.

  5. Add this exact OAuth redirect URI:

https://your-deployed-etsy-mcp.example.com/oauth/etsy/callback
  1. Replace the host with the real deployed host used in PUBLIC_BASE_URL.

  2. The URI must be HTTPS and must match exactly. No trailing slash.

The server requests these Etsy scopes:

shops_r shops_w listings_r listings_w transactions_r transactions_w

Local Development

npm install
cp .env.example .env
npm run dev

For local Agent Builder testing, expose the server with an HTTPS tunnel and set:

PUBLIC_BASE_URL=https://your-tunnel-host

Then add this redirect URI to Etsy:

https://your-tunnel-host/oauth/etsy/callback

Deployment

Deploy to any Node host that supports persistent storage for the JSON token store, such as Render, Railway, Fly.io, or a VPS.

Build command:

npm install && npm run build

Start command:

npm start

Set all environment variables in the hosting provider.

If your platform has ephemeral disk, set DATA_PATH to a mounted persistent volume path or replace the JSON store with Postgres/Redis before production use.

OpenAI Agent Builder Setup

In Agent Builder, add a remote MCP connection:

Field

Value

MCP Server URL

https://your-deployed-etsy-mcp.example.com

Authentication

OAuth

Headers

None

Do not use:

https://your-deployed-etsy-mcp.example.com/mcp
https://your-deployed-etsy-mcp.example.com/sse

When Agent Builder connects, it discovers:

https://your-deployed-etsy-mcp.example.com/.well-known/oauth-protected-resource
https://your-deployed-etsy-mcp.example.com/.well-known/oauth-authorization-server

The OAuth flow then sends Employee #1 to Etsy, Employee #1 approves the Etsy app, and Agent Builder receives an MCP bearer token for subsequent tool calls.

Etsy API Routes Used

The server calls Etsy Open API v3:

GET   /v3/application/users/{user_id}/shops
GET   /v3/application/shops/{shop_id}
GET   /v3/application/shops/{shop_id}/listings
GET   /v3/application/shops/{shop_id}/listings/active
POST  /v3/application/shops/{shop_id}/listings
PATCH /v3/application/shops/{shop_id}/listings/{listing_id}
GET   /v3/application/shops/{shop_id}/receipts

Every Etsy API call sends:

Authorization: Bearer <etsy_oauth_access_token>
x-api-key: <keystring>:<shared_secret>

Related MCP Connectors

  • The Stytch MCP server is a reference implementation that demonstrates remote MCP server authentication and authorization using Stytch Connected Apps. It provides OAuth 2.1-compliant authorization (including PKCE), Dynamic Client Registration, and validates Stytch-issued access tokens to enable AI agents to securely interact with external services through permissioned access, supporting scopes like openid, email, profile, and manage:project_data.

  • MCP server connecting AI agents to 100+ apps (Gmail, Slack, Notion, GitHub) via one-click OAuth.

  • The Remote MCP server acts as a standardized bridge between LLM applications (like Claude, ChatGPT, and Cursor) and external services, enabling AI agents to access external tools and resources. Its primary capability is providing a centralized search tool to discover other MCP servers and their respective tools. Unlike local implementations, it runs remotely with OAuth authentication and permission controls for security.

  • MCP server for secureFlows: token-free URL builders and integration-linting tools for AI agents.

Related MCP Servers

  • A
    license
    A
    quality
    D
    maintenance
    A production-grade MCP server that enables AI assistants like Claude to read Etsy shop data—listings, orders, inventory, and stats—through eight read-only tools, using OAuth and automatic pagination.
    8
    MIT
  • A
    license
    Not graded
    quality
    D
    maintenance
    MCP server for interacting with the Etsy API. Allows users to retrieve shop data, manage listings, and perform various Etsy operations through natural language.
    MIT