strategy-council

name	Strategy Council
description	Multi-perspective AI council for designing, validating, and improving trading strategies with 5 specialized agents.

Strategy Council Skill

A multi-agent discussion framework for rigorous strategy evaluation. Each "agent" represents a specialized perspective that challenges and validates trading strategies.

The Five Agents

1. The Quant 🧮 - Statistical rigor, Sharpe, p-values, sample size
2. The Risk Manager 🛡️ - Drawdown, position sizing, tail risk
3. The Execution Specialist ⚡ - Spreads, slippage, liquidity
4. The Regime Analyst 🌊 - Market conditions, when strategies work
5. The Devil's Advocate 😈 - Breaking strategies, finding weaknesses

How To Use

In Workflow (via /council):

Use the /council workflow to initiate a council session.

Programmatically:

from scripts.strategy_council import StrategyCouncil

council = StrategyCouncil()
result = council.evaluate_strategy(
    symbol="GOLD",
    strategy_type="breakout",
    strategy_params={"atr_mult": 2, "timeframe": "H1"}
)
print(result["verdict"])  # APPROVED, CONDITIONAL, REJECTED

Council Outputs

Each session produces:

• Individual agent analyses
• Combined synthesis
• Final verdict (APPROVED/CONDITIONAL/REJECTED)
• Action items for improvement

Integration

The council uses real market intelligence from:

• titan_system/core/comprehensive_intel.py - Live spread/ATR data
• config/alpha_registry.json - Historical validated edges
• data/market_intelligence_export.json - Symbol profiles