Bots

Create, configure and operate trading bots.

Creating a bot

SettingNotes
Name, wallet, symbols, initial capitalRequired.
Modeautonomous (default), dca, grid, momentum or mean_reversion.
AI provider / modelDefaults to the platform provider; pick one of your own LLMs instead if you prefer.
Refresh intervalSeconds between cycles. Default 300, minimum 30.
Leverage & margin typeSpot is spot; futures bots can set leverage. Locked after the bot's first trade.
Quote currencyDefault USDT.
Decide all symbolsLet the engine consider every eligible symbol for the quote currency.
Risk and mode configurationPosition size, stops and mode-specific parameters.
SupervisorEnabled and auto-apply by default, with an optional custom prompt (max 500 characters).

Lifecycle

  • Bots are created paused; press Start to activate.
  • While a bot is active you can change its name, instructions and risk, but indicators, symbols, mode configuration and refresh interval require a paused bot.
  • Kill all stops every active bot at once; open positions are not closed automatically.

Plan limits

The number of bots is set by your plan (see Plans & billing). Creating a bot over the limit returns 409 Bot limit reached.

Positions, blocked symbols and exports

  • Close now closes an open position at the latest price and records the trade.
  • Blocked symbols are never traded by any of your bots.
  • Export downloads all your trades as CSV (up to 10,000 rows).

How a cycle works

  1. Fetch klines from Binance market data and compute indicators.
  2. Detect the market regime (trending, ranging, high volatility, neutral).
  3. Run the veto pipeline: extreme volatility, low historical edge, timeframe misalignment, order-book depth.
  4. If a ThunderScript is active, evaluate entry when / exit when.
  5. Optionally ask the AI model for a final confidence check (ai_context).
  6. Execute, then record the decision, reasoning and snapshots in the audit trail.
AiTrade docs