Sam Trading AI agent

SAM (Sam Auto-Agent) is an autonomous AI trading agent built for the Alpaca Options Alpha Agents challenge. It combines a multi-agent LLM debate framework with Alpaca's Trading API, MCP server, and paper trading environment to autonomously analyze stocks, debate trading decisions, and execute options trades. HOW IT WORKS SAM runs a multi-agent graph with five layers: 1. Analyst team (market, social, news, fundamentals) — each analyst uses tool-calling to fetch real data via yfinance and Alpha Vantage. 2. Bull vs. bear debate — two researchers argue for and against the trade. 3. Research manager — judges the debate and produces an investment plan. 4. Trader — converts the plan into a BUY/SELL/HOLD proposal, options-aware. 5. Risk debate (risky/safe/neutral) + risk judge — final decision. The final decision is parsed and routed to AlpacaExecutor, which pulls the live option chain from Alpaca, picks the best contract based on direction + DTE + strike, sizes the position by 2% of equity, and submits a paper options order.