MCP Upstox Trading Server

What is MCP Upstox Trading Server?

An automated trading application that identifies and trades based on high-probability support and resistance levels (Most Connected Pivot) using historical price data. It is designed for traders using the Upstox brokerage platform.

How to use MCP Upstox Trading Server?

Install Node.js v14+, clone the project, run npm install, create a .env file with Upstox API credentials, then start the server with node server.js. Authenticate by visiting http://localhost:3000/auth.

Key features of MCP Upstox Trading Server

• Implements the Most Connected Pivot (MCP) technical analysis strategy
• Uses official Upstox JavaScript SDK for API calls
• Automatic token refresh and secure token persistence
• Supports multiple deployment options: local, Heroku, AWS EC2, Docker
• Provides REST endpoints for MCP calculation, trading, and automated strategy

Use cases of MCP Upstox Trading Server

• Automatically identify significant support/resistance levels for trading instruments
• Execute limit orders when price approaches identified MCP levels
• Run a fully automated MCP strategy with lookback and investment parameters
• Analyze multiple symbols simultaneously for MCP signals
• Backtest MCP strategy using historical data

FAQ from MCP Upstox Trading Server

What prerequisites are needed to run the server?

Node.js v14 or higher, NPM, an Upstox Developer Account with API access, and basic understanding of JavaScript and trading concepts.

How does the server handle authentication?

It stores access and refresh tokens in a local token.json file, automatically refreshes them 5 minutes before expiry, and provides a /logout endpoint to clear stored tokens.

What symbols and intervals does the server support?

Symbols must follow the Upstox format (e.g., NSE_FO|NIFTY-I). Intervals like 1D are supported, with configurable lookback dates via query parameters.

What deployment options are documented?

Local deployment with node server.js, cloud deployment on Heroku or AWS EC2 (using PM2), and containerized deployment via Docker.

How can the MCP calculation be customized?

By modifying the calculateMCP() function in server.js – adjusting connection criteria, pivot point identification, deviation thresholds, and trend identification logic.