TradeOS AI
@TradeOS-AI
Use TradeOS MCP Server to bring your trading data, indicators, strategy logic, and agent workflows into MCP-compatible AI tools. Copy your server URL, add it to GPT or Claude, and let your AI assistant work directly with your TradeOS trading intelligence.
概要
What is TradeOS AI?
TradeOS AI lets you connect TradeOS trading intelligence to MCP-compatible AI tools, so your assistant can analyze markets, search tickers, compare spreads, use macro context, and manage custom trading agents directly from natural-language prompts. It is a stdio MCP bridge that proxies tools to the TradeOS Streamable HTTP endpoint (https://ai.tradeos.xyz/api/agent/mcp/mcp-call).
How to use TradeOS AI?
Two modes are supported. HTTP (recommended): set the URL https://ai.tradeos.xyz/api/agent/mcp/mcp-call in your MCP client’s configuration (e.g., Cursor mcp.json); the client handles OAuth. npm stdio: install @tradeos/tradeos-mcp locally, run the OAuth helper to obtain a TRADEOS_ACCESS_TOKEN, then configure the client with the node command pointing to the bridge script and the token in env. Detailed steps for Cursor and Claude Code are provided in the README.
Key features of TradeOS AI
- Analyze markets and search tickers from natural language
- Compare spreads and use macro context
- Manage custom trading agents via prompts
- Supports both HTTP (OAuth) and npm stdio (token) transports
- Production endpoint:
https://ai.tradeos.xyz/api/agent/mcp/mcp-call - Compatible with Cursor, Claude Code, and most MCP clients
- Windows-friendly setup with local install instructions
Use cases of TradeOS AI
- Ask an AI assistant to analyze current market conditions
- Compare bid‑ask spreads across multiple instruments
- Let an AI agent manage a custom trading strategy
- Retrieve macro‑economic context to inform trading decisions
- Search tickers and fetch real‑time data via conversational prompts
FAQ from TradeOS AI
How do I get an access token for the npm stdio mode?
Run the OAuth helper: npx -y -p @tradeos/tradeos-mcp tradeos-mcp-oauth (or node …/scripts/fetch-token.mjs). It opens a browser to authenticate with TradeOS, then prints a JWT (eyJ…) to copy into your MCP client config.
What are the prerequisites for using TradeOS AI?
Node.js 18+ and a TradeOS account with MCP API access are required. For npm stdio mode, you need the ability to run a local installation.
Where does the data go when using TradeOS AI?
All tool calls are proxied to the production endpoint https://ai.tradeos.xyz/api/agent/mcp/mcp-call. No data is stored locally by the bridge; authentication is handled via browser OAuth (HTTP mode) or a token (npm stdio mode).
What transport and authentication options are available?
- HTTP (Streamable HTTP): browser OAuth managed by the MCP client.
- npm stdio: requires a
TRADEOS_ACCESS_TOKENenvironment variable (set in the client’senvconfig, not in the shell).
I see “TRADEOS_ACCESS_TOKEN is required” when connecting. What should I do?
The token is not set in the environment or mcp.json. Run the OAuth helper to get a fresh JWT, then paste it (without any Bearer prefix) into the env section of your MCP client configuration.
