Sakshi Raj@sakshi_raj

Options Alpha Agent is an autonomous options-trading system built on Alpaca's paper trading API. It splits decision-making into three independent stages. First, TradingAgents — an unmodified multi-agent LangGraph pipeline — debates each ticker across analyst and researcher roles to produce a BUY/SELL/HOLD signal with a confidence score. Second, a separate LLM structures the actual trade: strike, expiry, and structure (long calls/puts for high-conviction views, spreads when hedging fits better), grounded in live option-chain data it fetches itself via Alpaca's MCP server, sized within a disclosed risk budget — but with no ability to place an order. Third, twelve deterministic Python risk gates (max loss, aggregate exposure, liquidity, DTE bounds, circuit breakers) veto anything unsafe before an isolated executor places the trade through Alpaca's MCP server. Ticker selection is dynamic, not a fixed watchlist: each cycle screens real-time market movers and most-actives, stopping cleanly once the day's LLM quota is exhausted. Every decision, block, and trade is logged with its full audit trail on a live dashboard.

🌎 Online Hackathon | 💻 Dates: 28 August–4 September 2026 | 🏆 $6,000 Prize Pool Build AI trading agents on Alpaca — autonomous agents and trading apps using Alpaca's Trading API, MCP server and CLI. 🌍 Fully online hackathon. Join and build from anywhere in the world.