Compound Tools

Compound tools chain multiple operations into a single call. They cost less than calling each tool individually and provide a unified response.

pre_trade_gate

Cost: $0.01 | Complete pre-trade validation in one call.

Combines:

1. Position sizing (G-formula) to calculate the optimal number of shares.
2. Risk validation (Iron Fist rules) to approve or reject the trade.
3. System health (optional, if R-multiples provided) to assess current edge status.

Use this before every trade.

Parameters

Parameter	Type	Required	Description
symbol	string	Yes	Instrument symbol (e.g. "AAPL").
direction	string	Yes	"long" or "short".
entry_price	string	Yes	Planned entry price.
stop_price	string	Yes	Stop loss price.
equity	string	Yes	Account equity in USD.
daily_pnl	string	No	Daily P&L so far. Default "0".
risk_percent	string	No	Risk as decimal. Default "0.02" (2%).
r_multiples	string[]	No	Recent R-multiples. If provided, includes system health check.

Example

from systemr import SystemRClient

client = SystemRClient(api_key="sr_agent_...")

result = client.pre_trade_gate(
    symbol="AAPL",
    direction="long",
    entry_price="185.50",
    stop_price="180.00",
    equity="100000",
    r_multiples=["1.5", "-1.0", "2.3", "-0.5", "1.8", "-1.0", "3.2", "0.8"],
)

if result["gate_passed"]:
    shares = result["sizing"]["shares"]
    print(f"BUY {shares} shares of AAPL at $185.50, stop at $180.00")
else:
    print(f"Trade blocked: {result['risk']['errors']}")

Response

{
  "gate_passed": true,
  "sizing": {
    "shares": 363,
    "risk_amount": "2000.00",
    "risk_percent": "0.02",
    "notional": "67351.50",
    "one_r_dollars": "2000.00",
    "direction": "long"
  },
  "risk": {
    "approved": true,
    "score": 85,
    "errors": [],
    "warnings": [],
    "risk_amount": "2000.00",
    "risk_percent": "0.02"
  },
  "system_health": {
    "g": "0.0847",
    "expected_r": "0.7875",
    "verdict": "POSITIVE_EDGE",
    "trade_count": 8
  }
}

How gate_passed is determined

gate_passed is true when both conditions are met:

• sizing.shares > 0 (position size is valid)
• risk.approved == true (all Iron Fist rules pass)

System health is informational. A declining G does not block the gate. You can add your own logic to incorporate system health into your decision.

---

assess_trading_system

Cost: $2.00 | Comprehensive trading system assessment in a single call.

Chains 7 analyses together:

1. G-metric evaluation to measure geometric growth rate
2. Win/loss analysis for batting average and payoff ratio
3. Kelly criterion for optimal position sizing
4. Monte Carlo simulation to project future equity paths
5. Drawdown analysis to measure downside risk
6. What-if scenarios to find improvement opportunities
7. Variance killer detection to identify G-damaging trades

Parameters

Parameter	Type	Required	Description
r_multiples	string[]	Yes	R-multiples from trade history (10+ recommended).
equity_values	string[]	No	Equity values after each trade (for drawdown).
starting_equity	string	No	Starting equity. Default "100000".
pnl_values	string[]	No	USD P&L per trade (for detailed win/loss).

Example

result = client.assess_system(
    r_multiples=[
        "1.5", "-1.0", "2.3", "-0.5", "1.8", "-1.0", "3.2", "0.8",
        "-0.7", "1.1", "-0.3", "2.5", "-1.2", "1.6", "0.9",
    ],
    starting_equity="100000",
)

print(f"Verdict: {result['verdict']}")
print(f"G: {result['g_metrics']['g']}")
print(f"Win rate: {result['win_loss']['win_rate']}")
print(f"Kelly (half): {result['kelly']['half_kelly']}")
print(f"Monte Carlo profit probability: {result['monte_carlo']['probability_of_profit']}")
print(f"Max drawdown: {result['drawdown']['max_drawdown_pct']}")

Response structure

{
  "verdict": "POSITIVE_EDGE",
  "g_metrics": {
    "g": "0.0623",
    "expected_r": "0.62",
    "variance": "1.87",
    "trade_count": 15
  },
  "win_loss": {
    "win_rate": "0.667",
    "avg_win_r": "1.75",
    "avg_loss_r": "0.78",
    "payoff_ratio": "2.24",
    "profit_factor": "4.49",
    "expectancy": "0.91"
  },
  "kelly": {
    "full_kelly": "0.35",
    "half_kelly": "0.175",
    "quarter_kelly": "0.0875",
    "risk_of_ruin": "0.001"
  },
  "monte_carlo": {
    "probability_of_profit": "0.93",
    "probability_20pct_gain": "0.71",
    "probability_20pct_drawdown": "0.08",
    "median_final_equity": "118500"
  },
  "drawdown": {
    "max_drawdown_pct": "7.2",
    "longest_duration": 3,
    "recovery_factor": "4.8"
  },
  "what_if": {
    "cap_losses_at_1r": "0.0891",
    "avoid_worst_10pct": "0.0945"
  },
  "variance_killers": [
    {"index": 5, "r_multiple": "-1.2", "g_impact": "-0.012"}
  ]
}

Cost comparison

Calling all 7 tools individually would cost approximately $0.04. The compound tool costs $2.00 because it runs full-depth Monte Carlo (5,000 simulations) and comprehensive analysis at each step. Use individual tools when you need specific metrics; use assess_trading_system when you want the complete picture.

---

record_trade_journal

Cost: $0.003 | Record a completed trade to the agent’s journal.

Parameter	Type	Required	Description
symbol	string	Yes	Instrument symbol.
direction	string	Yes	"long" or "short".
entry_price	string	Yes	Entry price.
exit_price	string	Yes	Exit price.
stop_price	string	Yes	Stop loss price.
quantity	string	Yes	Position quantity.
r_multiple	string	No	Realized R-multiple.
pnl	string	No	Realized P&L in dollars.
trade_date	string	No	Trade date as ISO string (e.g. "2026-03-25").
notes	string	No	Free-text notes.

result = client.record_trade(
    symbol="AAPL",
    direction="long",
    entry_price="185.50",
    exit_price="192.00",
    stop_price="180.00",
    quantity="363",
    r_multiple="1.18",
    pnl="2360.50",
    trade_date="2026-03-25",
    notes="Breakout entry, rode to first target.",
)

Returns: confirmation with trade_id, symbol, r_multiple, pnl, recorded_at.