Ask Your Account Anything: 20 Prompts for Reviewing Your Own Trading

Ask Your Account Anything: 20 Prompts for Reviewing Your Own Trading

Most traders who connect an AI assistant to their MT5 account ask it one or two things, usually "how did I do this month," and then stop. That is a reasonable start, and it is exactly what MetaTrader 5 Now Talks to AI Assistants covers as the shorter introduction to this kind of connection. But a win rate and a monthly total are the least interesting things sitting in your trade history, compared with which setup you keep trading out of habit, whether your position size creeps up after a loss, or which two hour window quietly does most of the damage.

This is the expanded version of that idea: a full library of 20 prompts, grouped by what they actually investigate, ready to copy, paste, and run against your own connected account today. Every one is built as a question about what your account has already done, not a request for an opinion about what to do next. Nothing here asks an assistant to predict a market or recommend a trade, and none of the answers should be treated that way either.

How to Use This List

Every prompt below needs one thing to work: an AI assistant that can see your real trade history, not just discuss trading in the abstract. That means connecting an assistant such as Claude to your MT5 account through MCP, so it can query your actual closed trades and account balance rather than working from a summary you type from memory. If that connection is not set up yet, Connect Claude to MT5: Full Setup Guide walks through the client side configuration end to end.

Once connected, treat every prompt as a question about data that already exists. Swap in your own numbers, symbols, and date ranges wherever you see brackets, test on a demo account first if you have not connected a live one before, and read every answer as a description of your history, not as advice about your next trade.

Key Takeaways: • All 20 prompts below query your own closed trades, past sessions, and historical sizing. None of them ask an assistant to predict or recommend a future trade. • They are grouped into five areas: losses, position sizing, session timing, strategy consistency, and trade management, with four prompts in each. • Brackets like [symbol] or [X] percent mark where you fill in your own detail before running the prompt. • An assistant connected through MCP can read and summarize your history. Deciding what to do about what it finds is still entirely yours. • Run through a full list like this weekly or monthly rather than once. A single answer is only a snapshot. A repeated one shows an actual trend.

• Understanding your losses: Prompts: 4 | What it targets: Where money is actually leaking from the account
• Risk and position sizing: Prompts: 4 | What it targets: Whether sizing stays consistent or drifts under pressure
• Session and timing patterns: Prompts: 4 | What it targets: When your own results are strongest and weakest
• Strategy and setup consistency: Prompts: 4 | What it targets: Whether you are trading the plan you actually wrote down
• Trade management and exits: Prompts: 4 | What it targets: What happens to a trade after you click buy or sell

Understanding Your Losses

1. 1."Pull up my 20 largest losing trades from the past six months and group them by symbol, setup, and time of day. What pattern shows up most often?"

Reveals whether losses cluster around one instrument, setup, or time of day, rather than being spread evenly.

1. 1."Compare the average loss on trades I closed manually with the average loss on trades that hit my stop loss on their own. Which group cost more per trade?"

Reveals whether hesitating to accept a stop is turning small losses into larger ones.

1. 1."List every losing trade in the past year where the loss was more than double my average loss. What do those trades have in common?"

Reveals whether a few outsized losses are doing most of the damage to an otherwise reasonable record.

1. 1."How many of my losing trades closed within 5 minutes of opening, and how many stayed open for more than an hour? What was the average loss for each group?"

Reveals whether fast panic exits or slow "it will recover" exits are the costlier habit.

Risk and Position Sizing

1. 1."Calculate my position size as a percentage of account equity for every trade placed last quarter. How much did that percentage vary between my smallest and largest trade?"

Reveals whether sizing follows a consistent rule or drifts from trade to trade.

1. 1."Show every trade where my position size was above [X] percent of equity. What was the win rate and average result for that group compared with my normal size?"

Reveals whether larger trades have historically paid off, or just added risk.

1. 1."Look at the 10 trades I placed immediately after a losing trade. Did my position size go up, go down, or stay about the same compared with my average?"

Reveals a pattern of increasing size after a loss, sometimes called revenge sizing.

1. 1."What was my largest single day drawdown in the past year, and how many positions were open at the same time when it happened?"

Reveals correlated exposure across several positions that a per trade risk figure would not show.

Session and Timing Patterns

1. 1."Break my closed trades down by the hour they were opened. Which two hour window shows the highest total profit, and which shows the highest total loss?"

Reveals a personal timing edge, or blind spot, a simple win rate would never surface.

1. 1."Compare my results on trades placed in the 30 minutes after a major news release with everything else. What is the difference in outcome?"

Reveals whether trading straight into volatility spikes has historically helped or hurt.

1. 1."Group my trading history by day of the week. Is there a day where my results are consistently worse than the rest?"

Reveals schedule or fatigue related patterns tied to a specific day.

1. 1."How does my average holding time compare between profitable and losing trades on [symbol]?"

Reveals whether winners are cut short while losers are held too long, or the other way round.

Strategy and Setup Consistency

1. 1."Tag my last 50 trades by the setup or reason I noted at entry, where I recorded one. What share of trades had no reason logged at all?"

Reveals how much of your trading is planned versus impulsive.

1. 1."Of the trades where I noted a specific setup, which one has the highest win rate, and which has the worst average result?"

Reveals which parts of a strategy earn their place and which persist out of habit.

1. 1."Compare my results on trades that matched my written trading plan with trades that did not follow it. How large is the gap?"

Reveals the real cost, in account currency, of deviating from a plan you already wrote down.

1. 1."How consistent was my stop loss placement, in pips or in percentage terms, across trades on the same symbol?"

Reveals whether stops follow a repeatable rule or get set differently each time.

Trade Management and Exits

1. 1."For trades that were in profit at some point before closing at a loss or at breakeven, how much of that unrealized profit was given back on average?"

Reveals how much a missing exit plan costs, separate from the entry decision itself.

1. 1."Compare trades where I moved my stop loss after entry with trades where I left it untouched. What is the difference in outcome between the two groups?"

Reveals whether adjusting stops mid trade has historically helped or hurt.

1. 1."Look at every trade I closed early, before my original take profit or stop loss level was reached. What would the recorded outcome have been if the order had been left alone?"

Reveals the historical cost of early manual intervention, based purely on what already happened.

1. 1."Break down my exit reasons across the last 100 trades: stop loss hit, take profit hit, closed manually, or margin related. What share falls into each category?"

Reveals the balance between planned exits and reactive ones.

A Composite Example, Not a Real Account

A hypothetical trader running prompt 9 against six months of composite, illustrative history might get back something like "trades opened between 13:00 and 15:00 UTC show a combined loss, while trades opened between 07:00 and 09:00 UTC show a combined profit roughly three times larger, on a similar number of trades." That does not say anything about tomorrow's market. It says something true about what already happened during those two windows, which is the only kind of claim any of these 20 prompts are built to produce.

What This List Will Not Do

None of the prompts above ask an assistant to decide anything. A pattern in your last 100 trades is a fact about your account. What a live market will do next is not, and no assistant reading your MT5 history, however detailed, changes that.

LTI's own view on AI and MT5 together has not shifted: "The strongest trading setup is not the one with the most automation; it is the one you can understand, test, monitor, and control." The prompts in this list are built around that line. They ask an assistant to read, summarize, and count things that have already happened. Deciding what to do with that summary, trade by trade, is still entirely yours.

Risk disclosure: Trading in foreign exchange (forex), contracts for difference (CFDs), and other leveraged financial instruments involves a high degree of risk and may not be appropriate for all investors. You should only trade with funds that you can afford to lose without adversely affecting your financial position. No AI tool, including any assistant connected through MCP, is a guaranteed forex trader or capable of trading independently of you.

FAQ

Do I need to connect my live account to try these prompts?

No. Test the approach on a demo account first, the same way you would test any new tool connected to your terminal. There is no reason to expose a live account before you know an assistant is reading your data the way you expect.

Will the AI tell me what to trade next based on this history?

No, and it should not be asked to. Every prompt here describes what already happened in your account, not what to do next. Reviewing your last 50 trades can tell you a great deal about your own patterns. It cannot tell you what the market will do tomorrow.

What do I need set up before I can run these prompts?

An AI assistant connected to your MT5 trade history through MCP, so it can see your actual closed trades rather than a description you type in by hand. See Connect Claude to MT5: Full Setup Guide for the full setup, and MetaTrader 5 Now Talks to AI Assistants for the shorter introduction.

How often should I actually run through a list like this?

Weekly is a reasonable starting point, with a deeper monthly pass through everything above. A single answer only covers one week. The same prompt run every week for a couple of months starts showing an actual trend. If retyping the same prompts feels repetitive, How to Build an AI Trading Journal That Fills Itself covers a way to have most of this waiting for you automatically.

Next Steps

Twenty prompts is a lot to run through once and never reopen. Start with one group, probably Understanding Your Losses, and see what your account has been doing that a simple win rate never showed you. Add the rest as they become useful.

Ready to see what your own account has been telling you? [Connect your LTI account] to MT5, follow the setup guide above to link an AI assistant, and start with the first group on this page against your own trade history.