> ## Documentation Index > Fetch the complete documentation index at: https://docs.performax.ai/llms.txt > Use this file to discover all available pages before exploring further. # AI Chat > Ask questions about your own trading history. **Chat**, in the **AI** section of the sidebar, answers questions about your own Performax data. It is a review tool, not a trading system. The same AI also opens from a trade in the Journal, in its **AI Assistant** panel. For the daily briefing and your trader profile, see the [AI Center](/features/ai-center). **Not financial advice.** AI output is not investment, financial, legal, tax, or trading advice. It does not predict markets, generate signals, or tell you what to buy or sell. Performax demo AI Chat with a performance review conversation *Demo account. Verify AI text against Journal and trade details before using it in review.* ## What it can use The AI works from the data in your account. Trades on hidden accounts are left out. It can use: * your closed trades: date, pair, side, P\&L and exchange; * your dashboard and Analytics figures: win rate, profit factor, expectancy, average R, average win and loss, best and worst trade, max drawdown in dollars, losing streaks, total fees, Sharpe, Sortino and Calmar; * breakdowns by exchange, pair, side, hour of day and day of week (hours in your own time zone); * monthly P\&L, R distribution with MAE/MFE, holding time, and deposits and withdrawals; * your strategies: their rules and checklists, and the stats of one strategy at a time; * behavior over the last 30 days: quick re-entries after a loss, rapid-fire trades, size increases after losses, average leverage, funding paid, and your logged mood against your results; * your Coach: score, rules, violations and Autopsy; * your risk settings (risk per trade, daily loss limit, max drawdown, monthly goal), your golden rules, your Notebook and your calendar notes; * the market context: BTC volatility, funding rate and high-impact economic events in the next 3 days; * your **AI profile** settings (**Coach personality**, **Trading rhythm**) and the main weakness from your psychology profile, which set its tone. For open positions, it only knows how many there are. If the account has little data, treat the output as low-confidence review text. ## Asking questions **Chat** in the sidebar opens your conversations. Ask questions about your own data, for example: * `How did my BTC trades do over the last 30 days?` * `What is my win rate by pair?` * `Long vs short: where do I actually make money?` * `At what hours and on which days do I lose the most?` * `Show my losing streaks of 3 trades or more and what each one cost.` * `How much did fees and funding cost me?` * `Compare my last 6 months.` * `Did I break any of my Coach rules this week, and what did it cost me?` * `What did I write in my notebook about FOMO?` **Questions** in the message box offers ready-made questions. Good questions name a period, pair, strategy or behavior. Vague questions can produce vague answers. Replies are text, with lists and tables. Highlighted terms such as drawdown or R multiple show a short definition and **Ask AI about this term**. You can rename, pin, archive, delete and sort conversations into folders. A new conversation starts on its own when the last one is more than 24 hours old. On any reply, **Pin as rule** saves it to your golden rules in the [AI Center](/features/ai-center#golden-rules). The only change the AI can make for you is on the [Coach](/features/coach): turn on a rule or adjust its threshold when you ask. It does not turn a rule off and does not loosen a threshold by more than double. Do that yourself in Settings. ## Trade analysis To review one trade in detail, open it in the Journal, then open the **AI Assistant** panel. Click **Analyze this trade** or type your own question about that trade. This analysis reads the full trade record: entry and exit, size, leverage, P\&L with fees and funding, R multiple, stop and target, planned risk-reward, MAE/MFE, duration, linked strategy, mistake tags, mood, tags and notes. It reviews execution against your plan, sizing, any mistake and its cost, and the lesson to take away. This detail is only available from the Journal. In Chat, a typed question such as "analyze my last SOL loss" does not open this analysis. ## Behavior patterns The AI can describe patterns such as: * quick re-entries after a loss and what they cost; * rapid-fire trading; * position size increasing after losses; * high average leverage or heavy funding costs; * results that change with your logged mood; * performance differences by hour, day, pair, side or strategy; * losing streaks and their cost. Pattern text should include the sample it is based on when possible. Small samples are not reliable proof of an edge. ## Limits The AI cannot: * show charts or visual cards in its replies; * tell you what your Journal mistake tags cost (only the trade analysis reads them); * break your results down by leverage level or by market session; * give your current drawdown in percent or your monthly returns in percent; * read entry, exit, stop or leverage per trade in Chat; * take any action other than turning on or adjusting a Coach rule. In long conversations, it only takes the most recent 20 messages into account. A daily message limit applies. AI output can also be incomplete or wrong when trade history is missing, balances or deposits are incomplete, notes are ambiguous, the sample is too small, or the AI summarizes a pattern too confidently. Verify important claims against the Journal, Analytics, and raw trade details. This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.