Docs
Open Revolution St
Model Context Protocol

MCP server

Connect Claude, ChatGPT, Cursor and other AI apps to Revolution St.

Server URL (Streamable HTTP)
https://v3.revolutionstreetwxs.io/mcp
Authorization server
https://v3.revolutionstreetwxs.io/api/auth
Protected resource metadata (RFC 9728)
https://v3.revolutionstreetwxs.io/.well-known/oauth-protected-resource/mcp
Documentation server (no sign-in)
https://v3.revolutionstreetwxs.io/mcp/docs

Overview

Revolution St runs a Model Context Protocol server at https://v3.revolutionstreetwxs.io/mcp (Streamable HTTP). Any MCP client can connect to it. The client signs you in with OAuth using the normal Revolution St sign-in (password or SSO), and then has the same permissions as your role, in the production environment.

It has the same tools as the built-in assistant (data_describe, data_query, data_write, data_undo and the analytics tools), plus two documentation tools: wxs_describe_product (how a page, setting or warehouse term works) and wxs_describe_api (the REST API, endpoint by endpoint, with schemas).

Connecting a client

Setup steps for Claude, Claude Code, ChatGPT, Cursor, VS Code and Gemini CLI are below. Configuration → MCP Server in Revolution St shows the same steps with your server’s address filled in. In Claude Code:

claude mcp add --transport http wxs https://v3.revolutionstreetwxs.io/mcp

Sign in when the client asks. Clients register themselves (OAuth dynamic client registration), so there is no client ID to copy. Scripts can send an API key as the bearer token instead of signing in.

Documentation server

A second server at https://v3.revolutionstreetwxs.io/mcp/docs has only the two documentation tools. It needs no account or OAuth, is read-only and rate-limited per IP address, and cannot read any warehouse data. Use it to let a coding agent read the REST API reference on a machine without a Revolution St login, or share it with a partner.

For live data and actions, connect to https://v3.revolutionstreetwxs.io/mcp instead. Setup for each client is below.

Set up your client

Use the server address https://v3.revolutionstreetwxs.io/mcp. Each guide links to that app's documentation.

Claudeclaude.ai and the Claude desktop app
  1. In Claude, open Customize → Connectors (on Team and Enterprise plans an Owner adds it under Organization settings → Connectors first, then members see it).
  2. Click + and choose Add custom connector.
  3. Paste the server URL: https://v3.revolutionstreetwxs.io/mcp. Leave Advanced settings empty. No client ID or secret is needed.
  4. Click Add, then Connect. Your browser opens the Revolution St sign-in; sign in as usual (password or SSO) and click Allow.
  5. In a chat, open the + menu → Connectors and switch the connector on. Ask something like "how many orders are waiting for a wave?"

Free plans allow one custom connector; Pro, Max, Team and Enterprise are unlimited.

Claude documentation ↗

Claude CodeAnthropic's terminal agent
  1. In a terminal, add the server (once per machine with --scope user, or per project without it):
  2. claude mcp add --transport http --scope user wxs https://v3.revolutionstreetwxs.io/mcp
  3. Start claude, type /mcp, pick wxs and follow the browser sign-in. Tokens are stored by Claude Code and refreshed automatically.
  4. Try it: "use wxs to look up the API for creating orders and write me a Python script that imports a CSV".
Add the server
claude mcp add --transport http --scope user wxs https://v3.revolutionstreetwxs.io/mcp
Or check .mcp.json into a project
{
  "mcpServers": {
    "wxs": {
      "type": "http",
      "url": "https://v3.revolutionstreetwxs.io/mcp"
    }
  }
}

Claude Code documentation ↗

ChatGPTCustom connector (developer mode)
  1. In ChatGPT, open Settings and turn on Developer mode (under Apps & Connectors → Advanced settings, or Security and login, depending on your version).
  2. Open Settings → Apps & Connectors (or chatgpt.com/plugins), click + / Create.
  3. Enter a name (Revolution St) and the server URL https://v3.revolutionstreetwxs.io/mcp. Set Authentication to OAuth and leave the client ID and secret blank. ChatGPT registers itself.
  4. Click Create, then Connect and complete the Revolution St sign-in in the browser tab that opens.
  5. In a chat, turn on the connector from the + menu (Developer mode).

Custom connectors require a plan with developer mode (Plus, Pro, Business, Enterprise, Edu). Deep-research-only connectors need search/fetch tools; use developer mode for the full tool set.

ChatGPT documentation ↗

CursorAI code editor
  1. Open Cursor Settings → MCP (or the Customize page) and choose Add new MCP server, or edit the config file directly.
  2. Add the entry below to ~/.cursor/mcp.json (everywhere) or .cursor/mcp.json in a project.
  3. Back in Settings → MCP, click Connect/Login next to wxs; a browser tab opens the Revolution St sign-in. Allow access.
  4. Tools appear in the agent's tool list; toggle any you don't want.
~/.cursor/mcp.json
{
  "mcpServers": {
    "wxs": {
      "url": "https://v3.revolutionstreetwxs.io/mcp"
    }
  }
}

Cursor documentation ↗

VS CodeGitHub Copilot agent mode
  1. Open the Command Palette and run MCP: Add Server.
  2. Choose HTTP (HTTP or Server-Sent Events), paste https://v3.revolutionstreetwxs.io/mcp, name it wxs, and pick Global (every workspace) or Workspace (.vscode/mcp.json).
  3. Start the server from the MCP: List Servers command (or the play icon in mcp.json). VS Code opens your browser to sign in; Allow access.
  4. In Copilot Chat, switch to Agent mode and pick the wxs tools from the Tools menu.
.vscode/mcp.json
{
  "servers": {
    "wxs": {
      "type": "http",
      "url": "https://v3.revolutionstreetwxs.io/mcp"
    }
  }
}

VS Code documentation ↗

Gemini CLIGoogle's terminal agent
  1. Add the server from a terminal:
  2. gemini mcp add --transport http wxs https://v3.revolutionstreetwxs.io/mcp
  3. Start gemini and run /mcp auth wxs to sign in through the browser.
  4. Run /mcp to confirm the tools are listed.
Add the server
gemini mcp add --transport http wxs https://v3.revolutionstreetwxs.io/mcp
Or in ~/.gemini/settings.json
{
  "mcpServers": {
    "wxs": {
      "httpUrl": "https://v3.revolutionstreetwxs.io/mcp"
    }
  }
}

Gemini CLI documentation ↗

Any other MCP clientWindsurf, Zed, Copilot CLI, custom agents…
  1. Transport: Streamable HTTP at https://v3.revolutionstreetwxs.io/mcp (no SSE endpoint).
  2. Authorization: OAuth 2.1 authorization code with PKCE and Dynamic Client Registration. Clients that follow the MCP authorization spec find the endpoints from the server URL, so nothing else needs to be configured.
  3. Scripts and CI can send an API key from Configuration → API Access as Authorization: Bearer wxs_pk_… instead of signing in. The key's permissions apply.
  4. For documentation only, https://v3.revolutionstreetwxs.io/mcp/docs has the two wxs_describe_* tools and needs no authentication (read-only, rate-limited per IP address).
  5. To inspect the tools by hand: npx @modelcontextprotocol/inspector, choose Streamable HTTP, paste the URL and click Connect.
Discovery
curl https://v3.revolutionstreetwxs.io/.well-known/oauth-protected-resource/mcp
curl https://v3.revolutionstreetwxs.io/.well-known/oauth-authorization-server/api/auth
# Documentation server (no sign-in):
curl -X POST https://v3.revolutionstreetwxs.io/mcp/docs -H "Content-Type: application/json" -d '{"jsonrpc":"2.0","id":1,"method":"tools/list"}'
With an API key (no browser)
curl -X POST https://v3.revolutionstreetwxs.io/mcp \
  -H "Authorization: Bearer wxs_pk_…" \
  -H "Content-Type: application/json" -H "Accept: application/json" \
  -d '{"jsonrpc":"2.0","id":1,"method":"tools/list"}'

Any other MCP client documentation ↗

Documentation server setup

https://v3.revolutionstreetwxs.io/mcp/docs has only the two documentation tools and needs no sign-in. Add it to a coding agent to let it read the REST API reference.

Claude Code
claude mcp add --transport http --scope user wxs-docs https://v3.revolutionstreetwxs.io/mcp/docs
Cursor — ~/.cursor/mcp.json
{
  "mcpServers": {
    "wxs-docs": {
      "url": "https://v3.revolutionstreetwxs.io/mcp/docs"
    }
  }
}
VS Code — .vscode/mcp.json
{
  "servers": {
    "wxs-docs": {
      "type": "http",
      "url": "https://v3.revolutionstreetwxs.io/mcp/docs"
    }
  }
}
Try it from a terminal
curl -X POST https://v3.revolutionstreetwxs.io/mcp/docs \
  -H "Content-Type: application/json" -H "Accept: application/json" \
  -d '{"jsonrpc":"2.0","id":1,"method":"tools/list"}'

Tools

Tools marked as writes change data, and MCP apps ask before running them. The documentation tools are also on the documentation server.

Data

data_describeLists the entities you can read, or one entity's columns and enum values.
data_queryReads rows or grouped totals from any entity, with filters, sorting and paging.
data_writeCreates, updates or deletes records of an entity, one at a time or in batches.writes
data_undoUndoes one recent action on an order or a wave. Undoing an undo redoes it.writes

Analytics (activity logs)

analytics_describe_log_schemaLists the log tables and columns that analytics_query_logs can query.
analytics_query_logsRuns a read-only SQL query on the activity log tables.
analytics_query_datasetQueries a prepared dataset and shows the result as a chart.

Documentation

wxs_describe_productExplains how a page, setting or warehouse term works.
wxs_describe_apiDescribes the REST API: an overview, endpoint search, or one endpoint's schemas.
esc