You do not need to write a perfect prompt. Start with your idea, then add details like market, timeframe, indicators, entry rules, exit rules, and risk settings.
Describe your trading idea clearly
When asking QuantPilot to help with a strategy, explain the idea in plain language.
A useful strategy prompt usually includes:
the market you want to analyze
the timeframe or candle interval
the indicators or signals you want to use
when the strategy should enter a trade
when the strategy should exit a trade
how risk should be managed
whether you want research, a strategy draft, optimization, or validation
Weak prompt:
“Make me a profitable strategy.”
Better prompt:
“Create a long-only BTC strategy on Hyperliquid using 1-hour candles. Use RSI and moving averages to find pullbacks in an uptrend. Enter when RSI recovers from oversold and price is above the 200 EMA. Include stop loss, take profit, and basic risk controls.”
Include the market and timeframe
Tell QuantPilot which market and timeframe you want to work with.
Examples:
“Use BTC on Hyperliquid with 1-hour candles.”
“Analyze ETH using 15-minute candles.”
“Build a strategy for SOL using 4-hour candles.”
If you do not include a market or timeframe, QuantPilot may choose defaults or ask follow-up questions.
Describe your indicators and signals
If your idea uses indicators, include them in the prompt.
Examples:
RSI
EMA or SMA
MACD
Bollinger Bands
volume
price breakouts
trend filters
support and resistance rules
Weak prompt:
“Use indicators to find good entries.”
Better prompt:
“Use RSI, 50 EMA, and 200 EMA. Only look for long entries when the 50 EMA is above the 200 EMA and price is above both moving averages.”
Explain entry and exit rules
QuantPilot works better when you describe what should trigger an entry and what should close the trade.
Entry examples:
“Enter long when RSI crosses back above 30.”
“Enter short when price breaks below the previous day’s low.”
“Enter only when price is above the 200 EMA.”
Exit examples:
“Exit when RSI reaches 70.”
“Use a 2% stop loss and 4% take profit.”
“Close the trade if price closes below the 50 EMA.”
“Use a trailing stop after the trade moves 3% in profit.”
Weak prompt:
“Buy low and sell high.”
Better prompt:
“Enter long when RSI crosses above 30 after being below 30, only if price is above the 200 EMA. Exit when RSI reaches 70 or if price closes below the 50 EMA.”
Add risk rules
Risk rules help QuantPilot create a more complete strategy.
You can include:
stop loss size
take profit size
position sizing idea
maximum number of open trades
whether short trades are allowed
whether the strategy should avoid trading during high volatility
maximum drawdown preference
Example:
“Use a 2% stop loss, 4% take profit, and no more than one open trade at a time. Avoid short trades.”
Ask for research or strategy work
QuantPilot can help with different types of tasks. Be clear about what you want.
Use research prompts when you want to explore an idea before building.
Example: “Research whether RSI mean reversion works better on BTC 1-hour candles or 4-hour candles. Summarize the pros, risks, and what should be tested.”
Use strategy prompts when you want QuantPilot to create or edit a strategy.
Example: “Create a QuantScript strategy for BTC on Hyperliquid using RSI mean reversion with a 200 EMA trend filter.”
Use review prompts when you want feedback on an existing strategy.
Example: “Review this strategy and explain what could cause poor performance or overfitting.”
Ask for parameter optimization
Parameter optimization means testing different settings to see which ones perform better in a backtest.
Example prompts:
“Optimize the RSI length, oversold level, and take profit for this strategy.”
“Test RSI lengths from 7 to 21 and compare the results.”
“Find better stop loss and take profit settings, but avoid settings that only work on one short time period.”
“Optimize this strategy for BTC 1-hour candles and summarize which parameters are most stable.”
Good optimization prompts should explain what to optimize and what to avoid.
Better prompt:
“Optimize the RSI length, oversold level, stop loss, and take profit. Prioritize stable results across multiple time periods instead of the highest single backtest return.”
Ask for robustness validation
Robustness validation means checking whether a strategy may still work under different conditions, instead of only looking good in one backtest.
Example prompts:
“Test this strategy across different date ranges.”
“Check whether small parameter changes break the strategy.”
“Compare performance in trending and sideways markets.”
“Look for signs of overfitting.”
“Run robustness checks and summarize the main weaknesses.”
Better prompt:
“Validate this strategy for robustness. Test nearby parameter values, different date ranges, and different market conditions. Tell me whether the strategy looks stable or overfit.”
Examples of weak and strong prompts
Weak prompt:
“Find a good BTC strategy.”
Strong prompt:
“Research and create a BTC strategy on Hyperliquid using 1-hour candles. Focus on trend-following entries with EMA filters and RSI confirmation. Include entry rules, exit rules, stop loss, take profit, and explain the main risks before creating the strategy.”
Weak prompt:
“Improve this strategy.”
Strong prompt:
“Review this strategy and suggest improvements to reduce drawdown. Focus on stop loss, take profit, trade frequency, and filters that avoid sideways markets. Do not make the strategy more complex unless needed.”
Weak prompt:
“Optimize everything.”
Strong prompt:
“Optimize RSI length, EMA length, stop loss, and take profit. Compare the results across multiple date ranges and avoid choosing parameters only because they had the highest return in one backtest.”
Tips for better results
Be specific about what you want QuantPilot to do.
Include market, timeframe, and direction if you know them.
Explain your entry and exit idea in plain language.
Add risk rules early.
Ask QuantPilot to explain assumptions before building.
Ask for backtesting and robustness checks before trusting a result.
Review all generated strategy logic before using it.
Important
QuantPilot helps with research, strategy building, testing, and analysis. Backtest results are based on historical data and do not guarantee future performance.
Always review the strategy logic, risk settings, and backtest assumptions before using any strategy.