[Feature] The agents can't see what options traders see — add an Options Positioning Analyst

The problem

TradingAgents can tell me to BUY into earnings without knowing the options market is already pricing a ±9% move.

That's the first thing any real trader checks. The four current analysts (fundamentals, news, sentiment, technical) all look at what already happened. None of them answers the questions traders actually ask before a trade:

• Is the move already priced in? Implied move vs. historical realized move.
• Where will price get stuck or accelerate? The strikes with the heaviest call and put open interest.
• Will dealers amplify or dampen the move? Positive vs. negative gamma regime.
• Is someone quietly buying protection? Put skew and unusual volume vs. open interest.

What changes in the debate

Bull: Clean breakout into earnings. Options Analyst: The straddle already prices ±7.8% vs. a ±5.1% average realized move. Spot sits just under the biggest call-OI strike, and dealers are long gamma, so expect chop, not a squeeze.

(Illustrative numbers.) One paragraph like this changes the Trader's entry and the Risk team's sizing.

Proposal

An opt-in options_analyst, off by default so nothing changes for existing users or benchmarks.

• Deterministic math. All numbers are computed in Python (implied move, OI levels, GEX, skew, unusual activity). The LLM only interprets them, so it can't invent figures.
• No look-ahead. yfinance only serves today's options chain. On any backtest date the analyst returns the existing no-data sentinel instead of leaking the current chain. Historical backtests come later through a pluggable vendor.
• Honest assumptions. Public open interest doesn't show who is long or short, so the report states its GEX sign convention instead of presenting it as fact.
• Clean fallback. Tickers without listed options (non-US, crypto) are skipped gracefully.

Plan

1. Live-date analyst using yfinance, with the sentinel on historical dates.
2. A historical options-chain vendor so backtests can use the analyst.
3. An ablation study: same tickers and dates with vs. without the options analyst.

I build options-flow and dealer-positioning analytics (GEX/VEX) for US equities full-time and am happy to open the PR for step 1 if this fits the roadmap. Would you prefer a separate analyst, or a tool attached to the Market Analyst?