Simulation profile
paper_simThe whole paper-trading simulation, configurable. By default this engine is deliberately easy: a market order fills whole at the live price, maintenance margin is a flat 0.5%, the balance is a scorecard that never refuses an open, and there is no latency, no size impact and no partial fill. That is right for learning and optimistic for proving out a bot. This sets any of it: the book size your results are measured against and whether it is a real budget (margin_enforced), what crossing the book costs on entry and exit (slippage, slippage_bps, impact_bps_per_100k), how long your order is in flight (latency_ms), whether orders fill part-way (partial_fills, min_fill_pct), how often the venue refuses (reject_rate_pct), your own fee (taker_bps), funding (funding, funding_mult) and where liquidation sits (margin_venue, margin_tiers, mmr_pct). Easiest route: pass a preset - frictionless, realistic, binance, bybit, okx, hyperliquid, thin_book or brutal. Reading returns every field with its range, its default and what it changes, the presets, and what a fill would look like right now. Out-of-range values are refused, never clamped. Positions keep the simulation they were filled under. Requires an API key.
Tool call template
This tool declares no required parameters. Add optional parameters from the input schema as needed.
{
"id": 1,
"jsonrpc": "2.0",
"method": "tools/call",
"params": {
"arguments": {},
"name": "paper_sim"
}
}Input Schema
| Name | Required | Description | Default |
|---|---|---|---|
| reset | No | true = back to the engine defaults. | |
| preset | No | frictionless | realistic | binance | bybit | okx | hyperliquid | thin_book | brutal. A preset starts from the defaults, so it is a clean slate plus that configuration. Omit every field to READ. | |
| funding | No | false = never charge funding, which is how you separate an execution edge from a carry edge. | |
| mmr_pct | No | 0.05-10. Your own maintenance margin, beating margin_venue. | |
| slippage | No | Fill against yourself the way a real book does: 0.01% crypto majors, 0.05% other crypto, 0.005% forex, 0.02% stocks/metals/indices. | |
| taker_bps | No | 0-100. Your own per-side fee in basis points. 5.5 = 0.055%. | |
| latency_ms | No | 0-2000. Your order is in flight this long and the price is re-read afterwards. | |
| funding_mult | No | 0-5. Scale the real 8-hourly funding rate. | |
| leverage_max | No | Your own leverage ceiling below the market cap. 0 = the market cap. | |
| margin_tiers | No | Raise maintenance margin with size, as a risk-limit ladder does, so a large position liquidates sooner. | |
| margin_venue | No | binance | bybit | okx | bitget | mexc | kucoin | gate | kraken | hyperliquid | coinbase, or "marginpad". Moves ONLY where liquidation sits. | |
| min_fill_pct | No | 10-100. The worst fill ratio with partial_fills on. | |
| slippage_bps | No | 0-500. Your own crossing cost in basis points instead of that table. 10 = 0.10%. | |
| exit_slippage | No | Pay the crossing cost on the way out too (default true). A stop, take-profit or liquidation still fills at its level. | |
| partial_fills | No | Allow a market order to fill less than it asked for. | |
| margin_enforced | No | true = an open needing more than free margin is refused 402, as a venue refuses it. Free margin is balance minus committed margin; unrealized profit is not collateral, because every position is isolated. | |
| reject_rate_pct | No | 0-50. Share of opens refused 503 venue_rejected on purpose. | |
| max_notional_usd | No | Your own ceiling on one position’s notional. 0 = none. | |
| start_balance_usd | No | 100 to 10,000,000. The book return_pct, the equity curve and max drawdown are measured against. | |
| impact_bps_per_100k | No | 0-500. Extra basis points per $100,000 of notional, because size walks further down the book. |