explaingit

mgcrea/mcp-tastytrade

Analysis updated 2026-05-18

0TypeScriptAudience · developerComplexity · 3/5Setup · moderate

TLDR

An MCP server that lets Claude or other AI assistants view and manage a TastyTrade brokerage account.

Mindmap

mindmap
  root((mcp-tastytrade))
    What it does
      Connects Claude to TastyTrade
      Exposes account as AI tools
    Tech stack
      TypeScript
      Node.js
      WebSocket DXLink
      Docker
    Use cases
      Check balances and positions
      Get options greeks
      Opt-in trading
    Audience
      Developers
      Options traders
    Safety
      Trading opt-in
      Confirm flag required

Code map

Detail Auto

An interactive map of this repo's files and how they connect — its source is parsed live in your browser. Click Visualize to build it.

filefunction / class

What do people build with it?

USE CASE 1

Ask Claude to check account balances, open positions, or recent orders on a TastyTrade brokerage account.

USE CASE 2

Get live options greeks like delta and gamma directly through an AI assistant conversation.

USE CASE 3

Set up an AI-driven trading workflow with opt-in, confirmation-gated order placement.

USE CASE 4

Run a self-hosted MCP server via Docker to connect Claude Code to a TastyTrade sandbox account for testing.

What is it built with?

TypeScriptNode.jsWebSocketDocker

How does it compare?

mgcrea/mcp-tastytrade0xradioac7iv/tempfsabboskhonov/hermium
Stars000
LanguageTypeScriptTypeScriptTypeScript
Setup difficultymoderatemoderatemoderate
Complexity3/53/54/5
Audiencedeveloperdeveloperdeveloper

Figures from each repo's GitHub metadata at analysis time.

How do you get it running?

Difficulty · moderate Time to first run · 30min

Requires a TastyTrade OAuth client secret and refresh token from the developer portal.

In plain English

mcp-tastytrade is a TypeScript server that connects the TastyTrade brokerage to AI assistants like Claude Desktop or Claude Code via the Model Context Protocol, known as MCP. MCP is a standard way for AI tools to interact with external services, so once this server is running, you can ask Claude natural-language questions about your brokerage account and it will fetch live data to answer. The server exposes your TastyTrade account as a set of AI-readable tools: listing accounts and balances, viewing open and closed positions, browsing order history, searching instruments and market metrics, managing watchlists, and getting live price quotes. For options traders, it also provides greeks data, which are metrics like delta and gamma that describe how an option's price moves relative to the underlying asset. Live quotes come via DXLink, a WebSocket connection that the server keeps open as a persistent real-time channel and reuses across calls, returning repeat quote requests in under 50 milliseconds. A REST fallback mode is available for simpler deployments, though greeks are unavailable in that mode. Trading operations, meaning placing, cancelling, and replacing orders, are opt-in: you must explicitly enable them with an environment variable, and each order-placement call requires a confirm flag to prevent accidental submissions. A sandbox environment is supported for testing. Authentication uses OAuth2 with automatic token refresh on expiry. The tech stack is TypeScript on Node 22 and above, built with tsdown, tested with vitest, and available as a Docker image. It runs locally over stdio and integrates with Claude Code or Claude Desktop through a config file entry.

Copy-paste prompts

Prompt 1
Walk me through configuring mcp-tastytrade with my TastyTrade OAuth credentials.
Prompt 2
Explain how to enable trading tools safely with TASTYTRADE_ALLOW_TRADING and the confirm flag.
Prompt 3
Help me set up mcp-tastytrade in Claude Desktop using the Docker image instead of a local build.
Prompt 4
How does the DXLink connection in this server keep quote requests fast across repeated calls?

Frequently asked questions

What is mcp-tastytrade?

An MCP server that lets Claude or other AI assistants view and manage a TastyTrade brokerage account.

What language is mcp-tastytrade written in?

Mainly TypeScript. The stack also includes TypeScript, Node.js, WebSocket.

How hard is mcp-tastytrade to set up?

Setup difficulty is rated moderate, with roughly 30min to a first successful run.

Who is mcp-tastytrade for?

Mainly developer.

Open on GitHub → Explain another repo

This repo across BitVibe Labs

Verify against the repo before relying on details.