stonex-mcp
MCP server exposing StoneX Client REST API endpoints as tools, enabling account balances, positions, activity, and performance retrieval via natural language. Requires a user-supplied access token or session cookie.
README
stonex-mcp
Typed TypeScript client for the StoneX Client (torchapi) REST API, plus an MCP server that exposes those endpoints as tools.
Auth is not automated (no Okta/MFA in this project). You log into StoneX Client in a browser, export a cookies.txt file, and pass the path to that file on the CLI.
Requirements
- Bun 1.1+ (for local development /
preparebuild) - Node.js 18+ (to run the published
binbundle)
Install
bun install
Auth: cookies.txt
- Install a browser extension that exports Netscape
cookies.txt(e.g. Get cookies.txt LOCALLY). - Log into https://client.stonex.com/client/.
- Export cookies for
client.stonex.comto a file (e.g.client.stonex.com_cookies.txt). - Pass that path to the server:
stonex-mcp --cookies /path/to/client.stonex.com_cookies.txt
The server reads session cookies from that file and calls GET https://client.stonex.com/client/token only when it needs a Bearer JWT.
Token cache
Derived access tokens are cached on disk so repeated MCP tool calls do not mint a new token every time:
| OS | Path |
|---|---|
| Windows | %LOCALAPPDATA%\stonex-mcp\access-token.json |
| macOS | ~/Library/Caches/stonex-mcp/access-token.json |
| Linux | ${XDG_CACHE_HOME:-~/.cache}/stonex-mcp/access-token.json |
When the cached JWT is missing or near expiry, a new token is requested using the cookies file and written back to the cache. If refresh starts failing with 401, re-export cookies after logging into StoneX Client again.
Treat the cookies file and cached token like passwords. Do not commit them.
Run the MCP server
bun start -- --cookies /path/to/client.stonex.com_cookies.txt
# or after build:
node dist/stonex-mcp.js --cookies /path/to/client.stonex.com_cookies.txt
Cursor / Claude Desktop config
{
"mcpServers": {
"stonex": {
"command": "npx",
"args": [
"-y",
"stonex-mcp",
"--cookies",
"C:/Users/YOU/Downloads/client.stonex.com_cookies.txt"
]
}
}
}
npx runs the built Node bundle (dist/stonex-mcp.js, produced by bun run build / prepare). Local equivalent: "command": "bun", "args": ["run", "/path/to/stonex-mcp/src/cli.ts", "--cookies", "/path/to/cookies.txt"].
Typed client
import { StoneXClient } from "./src/client/index.ts";
import { sessionCookieFromCookiesTxtFile } from "./src/auth/cookiesTxt.ts";
import { resolveAuth } from "./src/auth/resolveAuth.ts";
import { writeTokenCache } from "./src/auth/tokenCache.ts";
const { accessToken, sessionCookie } = await resolveAuth(
"/path/to/client.stonex.com_cookies.txt",
);
const client = new StoneXClient({
accessToken,
sessionCookie,
onAccessToken: (token) => writeTokenCache(token),
});
const accounts = await client.getAccounts();
const summary = await client.getAccountSummary({
accountNumber: accounts[0]!.acctNo,
});
Or build a cookie header yourself with sessionCookieFromCookiesTxtFile(...).
Endpoints mirror the StoneX Client SPA (https://vulcan.stonex.com/torchapi): accounts, balances, summary, positions, activity, performance, projected income, documents, and related user/account metadata.
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.