What problem does it solve? Deciding whether a user request belongs with the strategy specialist and preparing a clean handoff is error-prone: the Hub might design strategies itself, fabricate missing parameters, or mangle the user's exact rule wording. This Skill defines the routing boundary and handoff contract so strategy, backtest, and optimization requests reach the strategy_analysis agent intact. ## Core Features & Use Cases - Intent Classification: Distinguishes strategy design, backtesting, optimization, robustness analysis, and strategy repair from ordinary stock analysis, prediction markets, and pure fundamentals work. - Universe Resolution: Resolves tradable tickers from user input or screener_lookup with supported filters only, defaulting to 20-30 tickers. - Structured Handoff: Populates the three strategy_analysis arguments (user_request verbatim, universe list, context bullets) without inventing parameters or restating user rules. - Terminal Output Relay: Recognizes completed, pending, and error responses from strategy_analysis and relays them verbatim without Hub-authored additions. - Use Case: A user asks to backtest a 5-minute RSI mean-reversion strategy on NVDA; the Skill routes the verbatim request with the resolved universe to strategy_analysis and relays the nine-section result directly. ## Quick Start Ask the assistant to backtest a trading strategy, for example: backtest a momentum strategy on large-cap tech stocks over the past year.
