ninjatrader-strategy-dev_skill

Overview

This skill provides NinjaTrader 8 strategy development patterns focused on high-performance trading with C# on .NET Framework 4.8. It highlights real-time price handling, unmanaged order control, memory-efficient coding, and rate-limited order modifications to keep live trading reliable and fast. Use it to avoid common real-time bugs and to harden strategies for production.

How this skill works

The guidance inspects common hot-paths (OnMarketData, OnBarUpdate, OnOrderUpdate) and prescribes efficient patterns: tick-level live price tracking, cached indicator values, circular buffers, and pooled StringBuilder logging. It also defines unmanaged order submission and stop/target management with rate-limiting and disconnect detection. The patterns include safety checks, error handling, and performance targets to validate behavior under live feed conditions.

When to use it

• Implementing or debugging NinjaScript strategies for live execution
• Replacing Close[0]-dependent logic with tick-level updates
• Building unmanaged order workflows and robust stop management
• Optimizing memory and CPU in high-frequency or long-running strategies
• Hardening strategies for broker/exchange rate limits and disconnects

Best practices

• Use OnMarketData (MarketDataType.Last) for live decisions; avoid Close[0] in hot paths
• Cache indicator results once per bar and reference cached values in OnMarketData
• Rate-limit order modifications (e.g., 1s minimum) to comply with exchange rules
• Use fixed-size collections or circular buffers; clean up dictionaries when orders complete
• Pool StringBuilder instances for logging to reduce allocations and GC pressure
• Validate stop/limit prices before submission and include null/bars checks

Example use cases

• Trailing stop updates using live ticks and cached ATR to avoid delayed bar-close updates
• Fully unmanaged entry + stop orders for manual control with SubmitOrderUnmanaged and ChangeOrder
• Disconnect handling that flattens positions after stale tick detection (e.g., >5s)
• Memory-stable long-running strategies using circular buffers and StringBuilder pooling
• High-frequency decision logic meeting OnMarketData <1ms by avoiding heavy computations in the tick path