AgentSkillsCN

paper-trade

当用户希望创建、起草或优化 NautilusTrader 策略实施时使用，例如通过“generate strategy code”、“build a trading strategy”、“create NautilusTrader strategy”或“prepare strategy parameters for optimization”等短语触发。

SKILL.md
--- frontmatter
name: paper-trade
description: Use when users ask to simulate live execution without real capital, including trigger phrases like "start paper trading", "run strategy in simulation", "monitor paper account", or "prepare for live deployment".

Paper Trade Skill

Workflow Checklist

Copy and track progress:

code
Paper Trading Progress:
- [ ] Step 1: Confirm backtest readiness and deployment scope
- [ ] Step 2: Configure paper trading environment
- [ ] Step 3: Start paper trading session with safeguards
- [ ] Step 4: Monitor positions, orders, and risk drift
- [ ] Step 5: Set alerts and review cadence
- [ ] Step 6: Evaluate readiness for live transition

Step 1: Confirm Backtest Readiness and Scope

Before paper trading, verify strategy already passed backtest review.

Required checks:

  • •Strategy file exists (strategies/<name>.py)
  • •Recent backtest metrics are acceptable (risk-adjusted returns and drawdown)
  • •User agrees on symbols, timeframe, and expected behavior

If no valid backtest exists, invoke skill with backtest first.

Step 2: Configure Paper Trading Environment

Use trading_ops to set simulation parameters:

  • •Paper account capital
  • •Allowed instruments and max position sizes
  • •Order type constraints
  • •Session schedule and market hours

Use financial_research for current market context (earnings/events) so the user knows what regime the paper run starts in.

Step 3: Start Paper Trading Session with Safeguards

Launch paper trading via trading_ops.

At launch, verify:

  • •Strategy loaded correctly
  • •Orders route to simulated environment only
  • •No live broker endpoint is active
  • •Initial risk limits are enforced

If any safety check fails, stop immediately and correct configuration before restart.

Step 4: Monitor Positions, Orders, and Risk Drift

Use trading_ops to monitor continuously:

  • •Open positions and exposure
  • •Pending/filled/cancelled orders
  • •Realized and unrealized PnL
  • •Rule violations (if any)

Use quant_analysis on paper-trade performance snapshots to check drift from backtest expectations:

  • •Volatility and drawdown profile
  • •Win/loss distribution changes
  • •Slippage sensitivity hints

Step 5: Set Alerts and Review Cadence

Configure practical alert thresholds through workflow orchestration:

  • •Max daily drawdown breach
  • •Position limit breach
  • •Abnormal order rejection burst
  • •Unexpected strategy inactivity

Define review cadence with user:

  • •Intraday checks for active systems
  • •Daily risk and trade quality review
  • •Weekly performance consistency review

Step 6: Evaluate Readiness for Live Transition

Summarize paper trading outcomes with explicit go/no-go criteria:

  • •Stability across multiple sessions
  • •Risk controls working as intended
  • •Behavior close to backtest assumptions

Safety warnings (required):

  • •Paper fills can be unrealistically favorable
  • •Live latency and slippage can degrade edge
  • •Psychological pressure differs with real capital

Do not transition to live automatically. Recommend invoking skill with live-trade only after user explicitly accepts residual risk and confirms readiness.