📟 @backtest-kit/cli

Zero-boilerplate CLI for launching backtests, paper trading, and live trading. Run any backtest-kit strategy from the command line — no setup code required.

Point the CLI at your strategy file, choose a mode, and it handles exchange connectivity, candle caching, UI dashboard, and Telegram notifications for you.

📚 Backtest Kit Docs | 🌟 GitHub

New to backtest-kit? The fastest way to get a real, production-ready setup is to clone the reference implementation — a fully working news-sentiment AI trading system with LLM forecasting, multi-timeframe data, and a documented February 2026 backtest. Start there instead of from scratch.

🏎️ CLI Init

Minimal scaffold — all boilerplate stays inside @backtest-kit/cli:

npx @backtest-kit/cli --init --output backtest-kit-project
cd backtest-kit-project
npm install
npm start -- --help

🤔 Philosophy

@backtest-kit/cli is designed to do two things well — and the same tool covers both.

1. The lightest possible runner for a solo quant on day one.

You write a strategy file, point the CLI at it, and you're trading. No DI container to learn, no project scaffold to fight, no infrastructure code to copy-paste. One developer, one strategy, one command:

npx @backtest-kit/cli --init
npx @backtest-kit/cli --backtest ./content/feb_2026.strategy/index.ts

That's the whole onboarding. The first day you have an idea, you can backtest it. The first week you have an edge, you can paper-trade it. The first month you have a P&L, you can run it live — same CLI, different flag.

2. Built-in monorepo tooling for when the business takes off.

The moment you start making money is the worst possible moment to rewrite your stack in another language. So the CLI is also a monorepo-grade runner from day one — even if you don't use it that way at first.

monorepo/
├── content/
│   ├── feb_2026.strategy/ 
│       ├── feb_2026.strategy.ts   # strategy production code
│       ├── feb_2026.test.ts       # developer playground
├── packages/
│   ├── shared-broker/             # shared broker code
│   ├── shared-signals/            # common indicators (RSI, MACD)

As a result: you used to backtest your first idea is the same tool you use to run a desk of strategies in production. No rewrite, no language switch, no framework migration when the business scales — only more files in the monorepo.

✨ Features

• 🚀 Zero Config: Run npx @backtest-kit/cli --backtest ./strategy.mjs — no boilerplate needed
• 🔄 Four Modes: Backtest on historical data, walker A/B comparison, paper trade on live prices, or deploy live bots
• 💾 Auto Candle Cache: Warms OHLCV cache for all required intervals before backtest starts
• 🌐 Web Dashboard: Launch @backtest-kit/ui with a single --ui flag
• 📬 Telegram Alerts: Send formatted trade notifications with charts via --telegram
• 🔌 Default Binance: CCXT Binance exchange schema registered automatically when none is provided
• 🧩 Module Hooks: Drop a live.module.mjs, paper.module.mjs, or backtest.module.mjs to register a Broker adapter. No manual wiring needed.
• 🗃️ Transactional Live Orders: Broker adapter intercepts every trade mutation before internal state changes — exchange rejection rolls back the operation atomically.
• 🔑 Pluggable Logger: Override the built-in logger with setLogger() from your strategy module
• 🛑 Graceful Shutdown: SIGINT stops the active run and cleans up all subscriptions safely

📋 What It Does

@backtest-kit/cli wraps the backtest-kit engine and resolves all scaffolding automatically:

| Mode | Command Line Args | Description | |------------------|----------------------------|----------------------------------------------| | Backtest | --backtest | Run strategy on historical candle data | | Walker | --walker | A/B compare multiple strategies on the same historical data | | Paper | --paper | Live prices, no real orders | | Live | --live | Real trades via exchange API | | UI Dashboard | --ui | Web dashboard at http://localhost:60050 | | Telegram | --telegram | Trade notifications with price charts | | PineScript | --pine | Run a local .pine indicator against exchange data | | Pine Editor | --editor | Open the visual Pine Script editor in the browser | | Candle Dump | --dump | Fetch and save raw OHLCV candles to a file | | PnL Debug | --pnldebug | Simulate per-minute PnL for a given entry price and direction | | Broker Debug | --brokerdebug | Fire a single broker commit against the live broker adapter | | Flush | --flush | Delete report/log/markdown/agent folders from strategy dump dir | | Init Project | --init | Scaffold a new backtest-kit project |

🚀 Installation

Add @backtest-kit/cli to your project and wire it up in package.json scripts:

npm install @backtest-kit/cli

{
  "scripts": {
    "backtest": "npx @backtest-kit/cli --backtest ./src/index.mjs",
    "paper":    "npx @backtest-kit/cli --paper    ./src/index.mjs",
    "start":    "npx @backtest-kit/cli --live     ./src/index.mjs"
  },
  "dependencies": {
    "@backtest-kit/cli": "latest",
    "backtest-kit": "latest",
    "ccxt": "latest"
  }
}

Or run once without installing:

npx @backtest-kit/cli --backtest ./src/index.mjs

📖 Quick Start

Create your strategy entry point (src/index.mjs). The file registers schemas via backtest-kit — @backtest-kit/cli is only the runner:

// src/index.mjs
import { addStrategySchema, addExchangeSchema, addFrameSchema } from 'backtest-kit';
import ccxt from 'ccxt';

// Register exchange
addExchangeSchema({
  exchangeName: 'binance',
  getCandles: async (symbol, interval, since, limit) => {
    const exchange = new ccxt.binance();
    const ohlcv = await exchange.fetchOHLCV(symbol, interval, since.getTime(), limit);
    return ohlcv.map(([timestamp, open, high, low, close, volume]) => ({
      timestamp, open, high, low, close, volume,
    }));
  },
  formatPrice: (symbol, price) => price.toFixed(2),
  formatQuantity: (symbol, quantity) => quantity.toFixed(8),
});

// Register frame (backtest only)
addFrameSchema({
  frameName: 'feb-2024',
  interval: '1m',
  startDate: new Date('2024-02-01'),
  endDate: new Date('2024-02-29'),
});

// Register strategy
addStrategySchema({
  strategyName: 'my-strategy',
  interval: '15m',
  getSignal: async (symbol) => {
    // return signal or null
    return null;
  },
});

Run a backtest:

npm run backtest -- --symbol BTCUSDT

Run with UI dashboard and Telegram:

npm run backtest -- --symbol BTCUSDT --ui --telegram

Run live trading:

npm start -- --symbol BTCUSDT --ui

🎛️ CLI Flags

| Command Line Args | Type | Description | |---------------------------|---------|--------------------------------------------------------------------| | --backtest | boolean | Run historical backtest (default: false) | | --walker | boolean | Run Walker A/B strategy comparison (default: false) | | --paper | boolean | Paper trading (live prices, no orders) (default: false) | | --live | boolean | Run live trading (default: false) | | --ui | boolean | Start web UI dashboard (default: false) | | --telegram | boolean | Enable Telegram notifications (default: false) | | --verbose | boolean | Log each candle fetch (default: false) | | --noCache | boolean | Skip candle cache warming before backtest (default: false) | | --noFlush | boolean | Skip removing report/log/markdown/agent folders before backtest run (default: false) | | --symbol | string | Trading pair (default: "BTCUSDT") | | --strategy | string | Strategy name (default: first registered) | | --exchange | string | Exchange name (default: first registered) | | --frame | string | Backtest frame name (default: first registered) | | --cacheInterval | string | Intervals to pre-cache before backtest (default: "1m, 15m, 30m, 4h") | | --brokerdebug | boolean | Fire a single broker commit against the live broker adapter (default: false) | | --commit | string | Commit type for --brokerdebug (default: "signal-open") |

Positional argument (required): path to your strategy entry point file (set once in package.json scripts).

{
  "scripts": {
    "backtest": "npx @backtest-kit/cli --backtest ./src/index.mjs"
  }
}

🏃 Execution Modes

Backtest

Runs the strategy against historical candle data using a registered FrameSchema.

{
  "scripts": {
    "backtest": "npx @backtest-kit/cli --backtest --symbol ETHUSDT --strategy my-strategy --exchange binance --frame feb-2024 --cacheInterval \"1m, 15m, 1h, 4h\" ./src/index.mjs"
  }
}

npm run backtest

Before running, the CLI removes the report, log, markdown, and agent folders from the strategy's dump/ directory, then warms the candle cache for every interval in --cacheInterval. On the next run, cached data is used directly — no API calls needed. Pass --noCache to skip cache warming, --noFlush to keep existing output folders.

Paper Trading

Connects to the live exchange but does not place real orders. Identical code path to live — safe for strategy validation.

{
  "scripts": {
    "paper": "npx @backtest-kit/cli --paper --symbol BTCUSDT ./src/index.mjs"
  }
}

npm run paper

Live Trading

Deploys a real trading bot. Requires exchange API keys configured in your .env or environment.

{
  "scripts": {
    "start": "npx @backtest-kit/cli --live --ui --telegram --symbol BTCUSDT ./src/index.mjs"
  }
}

npm start

Walker — A/B Strategy Comparison

Runs the same historical period against multiple strategy files and prints a ranked comparison report. Use it to pick the best variant before deploying to backtest or live.

{
  "scripts": {
    "walker": "npx @backtest-kit/cli --walker --symbol BTCUSDT --noCache ./content/feb_2026_v1.strategy.ts ./content/feb_2026_v2.strategy.ts ./content/feb_2026_v3.strategy.ts"
  }
}

npm run walker

Each positional argument is a separate strategy entry point. Before loading them, the CLI removes the report, log, markdown, and agent folders from each entry point's dump/ directory. Pass --noFlush to keep existing output. All files are loaded without changing process.cwd() — .env is read from the working directory only. After loading, addWalkerSchema is called automatically using the exchange and frame registered by the strategy files.

If no frame is registered, the CLI falls back to the last 31 days from Date.now() with a console warning.

Walker-specific flags:

| Flag | Type | Description | |------|------|-------------| | --walker | boolean | Enable Walker comparison mode | | --symbol | string | Trading pair (default: "BTCUSDT") | | --cacheInterval | string | Intervals to pre-cache (default: "1m, 15m, 30m, 4h") | | --noCache | boolean | Skip candle cache warming (default: false) | | --noFlush | boolean | Skip removing report/log/markdown/agent folders before walker run (default: false) | | --verbose | boolean | Log each candle fetch and strategy progress (default: false) | | --output | string | Output file base name (default: walker_{SYMBOL}_{TIMESTAMP}) | | --json | boolean | Save results as JSON to ./dump/<output>.json | | --markdown | boolean | Save report as Markdown to ./dump/<output>.md |

Output modes:

• No flag — print Markdown report to stdout
• --json — save Walker.getData() result as JSON and exit
• --markdown — save Walker.getReport() as .md file and exit

Module hook: ./modules/walker.module is loaded automatically before the comparison starts (same rules as other modes — .ts, .mjs, .cjs tried in order).

Example — compare three variants and save the report:

npx @backtest-kit/cli --walker \
  --symbol BTCUSDT \
  --noCache \
  --markdown \
  --output feb_2026_comparison \
  ./content/feb_2026_v1.strategy.ts \
  ./content/feb_2026_v2.strategy.ts \
  ./content/feb_2026_v3.strategy.ts
# → ./dump/feb_2026_comparison.md

🐙 Multiple Symbol Parallel

For Poweruser — skip unless needed. The standard flow runs one symbol from --symbol. Use --entry only to fan out one strategy across many symbols at once, or to drive *.background() from a UI / DB / API.

With --entry, the CLI does only the boilerplate — Setup, providers (--ui / --telegram), the matching ./modules/<mode>.module, SIGINT that stops every active run via *.list(), and shutdown() once listenDone* reports all your runs complete. Picking the symbol set, warming cache, and calling *.background() is on you.

The --entry flag is a modifier — combine it with exactly one of --backtest / --live / --paper / --walker. One positional: the path to your entry file.

npx @backtest-kit/cli --backtest --entry ./src/multi-symbol.mjs

Example: backtest a strategy on five symbols at once

// src/multi-symbol.mjs
import {
  addExchangeSchema,
  addFrameSchema,
  addStrategySchema,
  Backtest,
  warmCandles,
} from "backtest-kit";
import ccxt from "ccxt";

addExchangeSchema({
  exchangeName: "binance",
  getCandles: async (symbol, interval, since, limit) => {
    const exchange = new ccxt.binance();
    const ohlcv = await exchange.fetchOHLCV(symbol, interval, since.getTime(), limit);
    return ohlcv.map(([timestamp, open, high, low, close, volume]) => ({
      timestamp, open, high, low, close, volume,
    }));
  },
  formatPrice: (symbol, price) => price.toFixed(2),
  formatQuantity: (symbol, quantity) => quantity.toFixed(8),
});

addFrameSchema({
  frameName: "feb-2026",
  interval: "1m",
  startDate: new Date("2026-02-01"),
  endDate: new Date("2026-02-28"),
});

addStrategySchema({
  strategyName: "my-strategy",
  interval: "15m",
  getSignal: async (symbol) => null,
});

// Decide the symbol set yourself — from a UI, database, API, or just a list.
const symbols = ["BTCUSDT", "ETHUSDT", "SOLUSDT", "BNBUSDT", "XRPUSDT"];

for (const symbol of symbols) {

  // 
  // Optional
  // 
  // await warmCandles({
  //   exchangeName: "binance,
  //   from: new Date("2026-01-01T00:00:00Z"),
  //   to: new Date("2026-01-31T23:59:59Z"),
  //   interval: "1m",
  //   symbol,
  // })


  Backtest.background(symbol, {
    strategyName: "my-strategy",
    exchangeName: "binance",
    frameName: "feb-2026",
  });
}

The same shape works for --live --entry / --paper --entry (call Live.background() per symbol with your broker adapter)

🗂️ Monorepo Usage

@backtest-kit/cli works out of the box in a monorepo where each strategy lives in its own subdirectory. When the CLI loads your entry point file, it automatically changes the working directory to the file's location — so all relative paths (dump/, modules/, template/) resolve inside that strategy's folder, not the project root.

How It Works

Internally, ResolveService does the following before executing your entry point:

process.chdir(path.dirname(entryPoint))  // cwd → strategy directory
dotenv.config({ path: rootDir + '/.env' })            // load root .env first
dotenv.config({ path: strategyDir + '/.env', override: true })  // strategy .env overrides

Everything that follows — candle cache warming, report generation, module loading, template resolution — uses the new cwd automatically.

Project Structure

monorepo/
├── package.json              # root scripts (one per strategy)
├── .env                      # shared API keys (exchange, Telegram, etc.)
└── strategies/
    ├── oct_2025/
    │   ├── index.mjs             # entry point — registers exchange/frame/strategy schemas
    │   ├── .env                  # overrides root .env for this strategy 
    │   ├── modules (optional)
    │   |    ├── live.module.mjs       # broker adapter for --live mode (optional)
    │   |    ├── paper.module.mjs      # broker adapter for --paper mode (optional)
    │   |    ├── backtest.module.mjs   # broker adapter for --backtest mode (optional)
    │   ├── template/             # custom Mustache templates (optional)
    │   └── dump/                 # auto-created: candle cache + backtest reports
    └── dec_2025/
        ├── index.mjs
        ├── .env
        └── dump/

Root package.json

{
  "scripts": {
    "backtest:oct": "npx @backtest-kit/cli --backtest ./strategies/oct_2025/index.mjs",
    "backtest:dec": "npx @backtest-kit/cli --backtest ./strategies/dec_2025/index.mjs"
  },
  "dependencies": {
    "@backtest-kit/cli": "latest",
    "backtest-kit": "latest",
    "ccxt": "latest"
  }
}

npm run backtest:oct
npm run backtest:dec

Isolated Resources Per Strategy

| Resource | Path (relative to strategy dir) | Isolated | |--------------------------|-----------------------------------|------------------| | Candle cache | ./dump/data/candle/ | ✅ per-strategy | | Backtest reports | ./dump/ | ✅ per-strategy | | Broker module (live) | ./modules/live.module.mjs | ✅ per-strategy | | Broker module (paper) | ./modules/paper.module.mjs | ✅ per-strategy | | Broker module (backtest) | ./modules/backtest.module.mjs | ✅ per-strategy | | Config module (walker) | ./modules/walker.module.mjs | ✅ loaded once | | Telegram templates | ./template/*.mustache | ✅ per-strategy | | Environment variables | ./.env (overrides root) | ✅ per-strategy |

Each strategy run produces its own dump/ directory, making it straightforward to compare results across time periods — both by inspection and by pointing an AI agent at a specific strategy folder.

🔗 Shared Import Aliases

@backtest-kit/cli automatically turns every top-level folder in process.cwd() into a bare import alias available inside any strategy file. No configuration needed — just create the folder.

How It Works

When the CLI loads a strategy file, it scans the current working directory for subdirectories and registers each one as an import alias. The alias name is the folder name. Both barrel imports and deep subpath imports are supported:

| Import | Resolves to | |--------|-------------| | import { fn } from "utils" | <cwd>/utils/index.ts (or .js, .mjs, .cjs) | | import { calcRSI } from "math/rsi" | <cwd>/math/rsi.ts | | import { research } from "logic" | <cwd>/logic/index.ts | | import { ResearchResponseContract } from "logic/contract/ResearchResponse.contract" | <cwd>/logic/contract/ResearchResponse.contract.ts |

Project Structure

my-project/
├── utils/                    ← import { formatDate } from "utils"
│   └── index.ts
├── math/                     ← import { calcRSI } from "math/rsi"
│   └── rsi.ts
├── logic/                    ← import { research } from "logic"
│   ├── index.ts              ←   barrel
│   └── contract/
│       └── ResearchResponse.contract.ts  ← import { ... } from "logic/contract/ResearchResponse.contract"
└── content/
    ├── feb_2026.strategy.ts  ← uses all three aliases freely
    └── mar_2026.strategy.ts  ← same aliases, no duplication

This lets you extract shared utilities, math helpers, or AI agent logic (e.g. agent-swarm-kit workflows) into named folders and reuse them across every strategy in the project without relative path hell.

TypeScript Support

Add a matching paths entry to your tsconfig.json so the editor resolves the aliases:

{
  "compilerOptions": {
    "moduleResolution": "bundler",
    "paths": {
      "logic": ["./logic/index.ts"],
      "logic/*": ["./logic/*"],
      "math": ["./math/index.ts"],
      "math/*": ["./math/*"],
      "utils": ["./utils/index.ts"],
      "utils/*": ["./utils/*"]
    }
  },
  "include": [
    "./logic",
    "./math",
    "./utils",
    "./content",
    "./modules",
  ],
}

🔔 Integrations

Web Dashboard (--ui)

Starts @backtest-kit/ui server. Access the interactive dashboard at:

http://localhost:60050

Customize host/port via environment variables CC_WWWROOT_HOST and CC_WWWROOT_PORT.

Symbol List (symbol.config)

By default the UI shows all symbols from the exchange. To restrict or reorder the list, create a config/symbol.config file in your strategy directory (next to the entry point).

Resolution order — first match wins:

| Priority | Path | Notes | |----------|------|-------| | 1 | {strategyDir}/config/symbol.config | per-strategy override (cwd after chdir) | | 2 | {projectRoot}/config/symbol.config | project-root override (cwd where npx was invoked) | | 3 | @backtest-kit/cli/config/symbol.config | built-in default shipped with the package |

Supported file formats (.ts, .cjs, .mjs, .js tried automatically):

// config/symbol.config.ts
export const symbol_list = [
  {
    icon: "/icon/btc.png",
    logo: "/icon/128/btc.png",
    symbol: "BTCUSDT",
    displayName: "Bitcoin",
    color: "#F7931A",
    priority: 50,
    description: "Bitcoin - the first and most popular cryptocurrency",
  },
  {
    icon: "/icon/eth.png",
    logo: "/icon/128/eth.png",
    symbol: "ETHUSDT",
    displayName: "Ethereum",
    color: "#6F42C1",
    priority: 50,
    description: "Ethereum - a blockchain platform for smart contracts",
  },
];

Notification Filter (notification.config)

Controls which notification categories are shown in the UI dashboard. Create a config/notification.config file in your strategy directory to override the defaults.

Resolution order — first match wins:

| Priority | Path | Notes | |----------|------|-------| | 1 | {strategyDir}/config/notification.config | per-strategy override (cwd after chdir) | | 2 | {projectRoot}/config/notification.config | project-root override (cwd where npx was invoked) | | 3 | @backtest-kit/cli/config/notification.config | built-in default shipped with the package |

Default values (built-in):

| Key | Default | Description | |-----|---------|-------------| | signal | true | Signal lifecycle: opened, scheduled, closed, cancelled | | risk | true | Risk manager rejection notifications | | info | true | Informational messages attached to an active signal | | breakeven | true | Breakeven level reached | | common_error | true | Non-fatal runtime errors | | critical_error | true | Fatal errors that terminate the session | | validation_error | true | Strategy config / input validation errors | | strategy_commit | true | All committed actions (partial close, DCA, trailing, etc.) | | partial_loss | false | Partial loss level reached (before commit) | | partial_profit | false | Partial profit level reached (before commit) | | signal_sync | false | Live order fill / exit confirmations from exchange sync |

// config/notification.config.ts
export default {
  signal: true,
  risk: true,
  info: true,
  breakeven: true,
  common_error: true,
  critical_error: true,
  validation_error: true,
  strategy_commit: true,
  partial_loss: false,
  partial_profit: false,
  signal_sync: false,
};

Telegram Notifications (--telegram)

Sends formatted HTML messages with 1m / 15m / 1h price charts to your Telegram channel for every position event: opened, closed, scheduled, cancelled, risk rejection, partial profit/loss, trailing stop/take, and breakeven.

Requires CC_TELEGRAM_TOKEN and CC_TELEGRAM_CHANNEL in your environment.

Telegram Message Adapter (telegram.config)

By default messages are rendered from Mustache templates (template/*.mustache). To override rendering programmatically, create a config/telegram.config file and export an object with any subset of get*Markdown methods. Each method receives the event payload and must return a Promise<string> with the Markdown message body.

Resolution order is the same as other configs (strategy dir → project root → package default).

// config/telegram.config.ts
import {
  IStrategyTickResultOpened,
  IStrategyTickResultClosed,
  RiskContract,
} from "backtest-kit";

export default {
  async getOpenedMarkdown(event: IStrategyTickResultOpened): Promise<string> {
    return `**Opened** ${event.symbol} at ${event.priceOpen}`;
  },
  async getClosedMarkdown(event: IStrategyTickResultClosed): Promise<string> {
    return `**Closed** ${event.symbol} at ${event.priceClosed}`;
  },
  async getRiskMarkdown(event: RiskContract): Promise<string> {
    return `**Risk rejected** ${event.symbol}`;
  },
};

All methods are optional — unimplemented ones fall back to the Mustache template.

| Method | Event type | |--------|------------| | getOpenedMarkdown | IStrategyTickResultOpened | | getClosedMarkdown | IStrategyTickResultClosed | | getScheduledMarkdown | IStrategyTickResultScheduled | | getCancelledMarkdown | IStrategyTickResultCancelled | | getRiskMarkdown | RiskContract | | getPartialProfitMarkdown | PartialProfitCommit | | getPartialLossMarkdown | PartialLossCommit | | getBreakevenMarkdown | BreakevenCommit | | getTrailingTakeMarkdown | TrailingTakeCommit | | getTrailingStopMarkdown | TrailingStopCommit | | getAverageBuyMarkdown | AverageBuyCommit | | getSignalOpenMarkdown | SignalOpenContract | | getSignalCloseMarkdown | SignalCloseContract | | getCancelScheduledMarkdown | CancelScheduledCommit | | getClosePendingMarkdown | ClosePendingCommit | | getSignalInfoMarkdown | SignalInfoContract |

🧩 Module Hooks (Broker Adapter)

The CLI supports mode-specific module files that are loaded as side-effect imports before the strategy starts. Each file is expected to call Broker.useBrokerAdapter() from backtest-kit to register a broker adapter.

| Command Line Args | Module file | Loaded before | |-------------------|---------------------------------|-----------------------------| | --live | ./modules/live.module.mjs | Live.background() | | --paper | ./modules/paper.module.mjs | Live.background() (paper) | | --backtest | ./modules/backtest.module.mjs | Backtest.background() | | --walker | ./modules/walker.module.mjs | Walker.background() | | --brokerdebug | ./modules/brokerdebug.module.mjs | broker commit test |

File is resolved relative to cwd (the strategy directory). All of .mjs, .cjs, .ts extensions are tried automatically. Missing module is a soft warning — not an error.

How It Works

The module file is a side-effect import. When the CLI loads it, your code runs and registers the adapter. From that point on, backtest-kit intercepts every trade-mutating call through the adapter before updating internal state — if the adapter throws, the position state is never changed.

// live.module.mjs
import { Broker } from 'backtest-kit';
import { myExchange } from './exchange.mjs';

class MyBroker {
  async onSignalOpenCommit({ symbol, priceOpen, direction }) {
    await myExchange.openPosition(symbol, direction, priceOpen);
  }

  async onSignalCloseCommit({ symbol, priceClosed }) {
    await myExchange.closePosition(symbol, priceClosed);
  }

  async onPartialProfitCommit({ symbol, cost, currentPrice }) {
    await myExchange.createOrder({
      symbol,
      side: 'sell',
      quantity: cost / currentPrice,
    });
  }

  async onAverageBuyCommit({ symbol, cost, currentPrice }) {
    await myExchange.createOrder({
      symbol,
      side: 'buy',
      quantity: cost / currentPrice,
    });
  }
}

Broker.useBrokerAdapter(MyBroker);

Broker.enable();

Available Broker Hooks

| Method | Payload type | Triggered on | |--------------------------|------------------------------|---------------------------| | onSignalOpenCommit | BrokerSignalOpenPayload | Position activation | | onSignalCloseCommit | BrokerSignalClosePayload | SL / TP / manual close | | onPartialProfitCommit | BrokerPartialProfitPayload | PP | | onPartialLossCommit | BrokerPartialLossPayload | PL | | onTrailingStopCommit | BrokerTrailingStopPayload | SL adjustment | | onTrailingTakeCommit | BrokerTrailingTakePayload | TP adjustment | | onBreakevenCommit | BrokerBreakevenPayload | SL moved to entry | | onAverageBuyCommit | BrokerAverageBuyPayload | DCA entry |

All methods are optional. Unimplemented hooks are silently skipped. In backtest mode all broker calls are skipped automatically — no adapter code runs during backtests.

TypeScript

import { Broker, IBroker, BrokerSignalOpenPayload, BrokerSignalClosePayload } from 'backtest-kit';

class MyBroker implements Partial<IBroker> {
  async onSignalOpenCommit(payload: BrokerSignalOpenPayload) {
    // place open order on exchange
  }

  async onSignalCloseCommit(payload: BrokerSignalClosePayload) {
    // place close order on exchange
  }
}

Broker.useBrokerAdapter(MyBroker);

Broker.enable();

⚙️ Setup Hook (config/setup.config)

@backtest-kit/cli loads a {projectRoot}/config/setup.config file once before any module hooks or strategy code run. Use it to perform one-time initialization that must happen before the first persistence call — registering a custom storage backend, configuring a logger, seeding global state, or anything else the process needs before backtest-kit starts.

Important: When setup.config is present, the CLI skips its own default adapter registration — it does not call usePersist() / useLocal() / useMemory() for any of the persistence slots. This means your config takes full ownership of the persistence layer: whatever adapters you register in {projectRoot}/config/setup.config are the ones backtest-kit uses, with no interference from the CLI defaults.

Example: MongoDB + Redis persistence via @backtest-kit/mongo

The most common use-case is swapping the default file-based persistence for a production-grade backend. Install @backtest-kit/mongo and call setup() — it registers all 15 persistence adapters in one call and reads connection parameters from environment variables:

npm install @backtest-kit/mongo

// config/setup.config.ts
import { setup } from '@backtest-kit/mongo';

setup();

# .env
CC_MONGO_CONNECTION_STRING=mongodb://localhost:27017/backtest-kit
CC_REDIS_HOST=127.0.0.1
CC_REDIS_PORT=6379

Or pass connection parameters explicitly:

// config/setup.config.ts
import { setup } from '@backtest-kit/mongo';

setup({
  CC_MONGO_CONNECTION_STRING: 'mongodb://mongo:27017/mydb',
  CC_REDIS_HOST: 'redis',
  CC_REDIS_PORT: 6379,
  CC_REDIS_PASSWORD: 'secret',
});

No changes to strategy code are needed — setup() wires up the adapters transparently before backtest-kit makes its first persistence call.

🧩 Module Loader (config/loader.config)

@backtest-kit/cli loads a {projectRoot}/config/loader.config file after setup.config but before any strategy or module code runs. Unlike setup.config (which is loaded for its side effects), loader.config exports a function that the CLI explicitly awaits. Use it whenever you need to wait for an async dependency to be ready before the backtest starts.

monorepo/
├── packages/
│   ├── shared-broker/      # shared broker code
│   ├── shared-signals/     # common indicators (RSI, MACD)
│   ├── shared-db/          # mongodb wiring
│   ├── strategy-momentum/  # strategy code
│   └── strategy-mean-reversion/

When to use it:

• Wire microfrontends in a monorepo — resolve and pre-load sibling packages, register cross-package services, or hydrate a shared DI container before strategies import from neighboring workspaces.
• Wait for a database connection — open a Mongo/Postgres/Redis connection and verify it's reachable before the first persistence call, so the backtest fails fast instead of mid-run.
• Warm up caches or external APIs — pre-fetch reference data (instruments list, calendar, fee tables) so the strategy's first tick doesn't pay the round-trip cost.
• Run schema migrations — apply any pending migrations to the persistence backend before signals start flowing.

loader.config supports exactly one of two export styles — never both at once. If both are present, the default export wins and the named loader is ignored.

// config/loader.config.ts — default export (preferred, ESM style)
export default async () => {
  await mongoose.connect(process.env.CC_MONGO_CONNECTION_STRING!);
  await redis.ping();
};

// config/loader.config.ts — named export
export const loader = async () => {
  await mongoose.connect(process.env.CC_MONGO_CONNECTION_STRING!);
  await redis.ping();
};

Example: wait for MongoDB before running a backtest

@backtest-kit/mongo's setup() registers the adapters synchronously but doesn't block until the connection is established. If your backtest depends on data that must be present in Mongo before the first signal fires, use loader.config to gate the run on a real connection:

// config/setup.config.ts
import { setup } from '@backtest-kit/mongo';

setup();

// config/loader.config.ts
import mongoose from 'mongoose';

export default async () => {
  await mongoose.connect(process.env.CC_MONGO_CONNECTION_STRING!);
  console.log('mongo connection verified, starting backtest');
};

Example: stitch microfrontends in a monorepo

When backtest-kit strategies live in one workspace and shared services (broker adapters, signal feeds, dashboards) live in sibling workspaces, loader.config is the place to wire them together before the runner starts:

// config/loader.config.ts
import "@my-org/brokers";
import "@my-org/signals";

The @my-org alias should be declared in config/alias.config.

🔀 Import Aliases (config/alias.config)

@backtest-kit/cli lets you override any nodejs module import — without touching the strategy code. Drop a config/alias.config file in your project root and export a mapping from module name to replacement module.

The alias table is loaded once (on the first import call) from {projectRoot}/config/alias.config and applied globally to every subsequent module load via require/ import.

Use cases:

• Replace a heavy dependency with a lighter stub for backtesting
• Swap any external api for a mock during CI runs

// config/alias.config.ts — named export
export const ccxt = require("./stubs/ccxt.stub.cjs");

// config/alias.config.cjs — default export
module.exports = {
  ccxt: require("./stubs/ccxt.stub.cjs"),
};

// config/alias.config.mjs — default export
import ccxtStub from "./stubs/ccxt.stub.mjs";

export default {
  ccxt: ccxtStub,
};

How it works: when strategy code calls require("ccxt"), the loader checks IMPORT_ALIAS first. If a key matches, the mapped value is returned instead of the real module — no monkey-patching of node_modules needed.

Important: It is not per-strategy — it applies to all modules loaded in the current process.

📦 Supported Entry Point Formats

@backtest-kit/cli automatically detects the format of your strategy file and loads it with the appropriate runtime — no flags or configuration required.

| Format | Extension | Runtime | Use Case | |--------|-----------|---------|----------| | TypeScript | .ts | tsx via tsImport() | TypeScript strategies with cross-imports (ESM ↔ CJS) | | ES Module | .mjs | Native import() | Modern JavaScript with top-level await and ESM syntax | | CommonJS | .cjs | Native require() | Legacy or dual-package strategies |

TypeScript (.ts)

Run TypeScript strategy files directly — no tsc compilation step needed. Powered by tsx, which handles cross-format imports transparently:

{
  "scripts": {
    "backtest": "npx @backtest-kit/cli --backtest ./src/index.ts"
  },
  "dependencies": {
    "@backtest-kit/cli": "latest",
    "backtest-kit": "latest",
    "tsx": "latest"
  }
}

ES Module (.mjs)

Standard ESM format. Supports top-level await, named exports, and import syntax:

{
  "scripts": {
    "backtest": "npx @backtest-kit/cli --backtest ./src/index.mjs"
  }
}

CommonJS (.cjs)

For projects that compile to or use CommonJS. Loaded via require():

{
  "scripts": {
    "backtest": "npx @backtest-kit/cli --backtest ./dist/index.cjs"
  }
}

🌲 Running Local PineScript Indicators

@backtest-kit/cli can execute any local .pine file against a real exchange and print the results as a Markdown table — no TradingView account required.

CLI Flags

| Flag | Type | Description | |------|------|-------------| | --pine | boolean | Enable PineScript execution mode | | --symbol | string | Trading pair (default: "BTCUSDT") | | --timeframe | string | Candle interval (default: "15m") | | --limit | string | Number of candles to fetch (default: 250) | | --when | string | End date for candle window — ISO 8601 or Unix ms (default: now) | | --exchange | string | Exchange name (default: first registered, falls back to CCXT Binance) | | --output | string | Output file base name without extension (default: .pine file name) | | --json | boolean | Write plots as a JSON array to <pine-dir>/dump/{output}.json | | --jsonl | boolean | Write plots as JSONL (one row per line) to <pine-dir>/dump/{output}.jsonl | | --markdown | boolean | Write Markdown table to <pine-dir>/dump/{output}.md |

Important: limit must cover indicator warmup bars — rows before warmup completes will show N/A

Positional argument: path to the .pine file.

Exchange via pine.module

By default the CLI registers CCXT Binance automatically. To use a different exchange — or to configure API keys, custom rate limits, or a non-spot market — create a modules/pine.module.ts file. The CLI loads it automatically before running the script.

The CLI looks for modules/pine.module in two locations (first match wins):

1. Next to the .pine file — <pine-file-dir>/modules/pine.module.ts
2. Project root — <cwd>/modules/pine.module.ts

my-project/
├── math/
│   ├── impulse_trend_15m.pine         ← indicator
│   └── modules/
│       └── pine.module.ts            ← loaded first (next to .pine file)
├── modules/
│   └── pine.module.ts                ← fallback (project root)
└── package.json

Inside pine.module.ts call addExchangeSchema from backtest-kit and give the exchange a name:

// modules/pine.module.ts
import { addExchangeSchema } from "backtest-kit";
import ccxt from "ccxt";

addExchangeSchema({
  exchangeName: "my-exchange",
  getCandles: async (symbol, interval, since, limit) => {
    const exchange = new ccxt.bybit({ enableRateLimit: true });
    const ohlcv = await exchange.fetchOHLCV(symbol, interval, since.getTime(), limit);
    return ohlcv.map(([timestamp, open, high, low, close, volume]) => ({
      timestamp, open, high, low, close, volume,
    }));
  },
  formatPrice: (symbol, price) => price.toFixed(2),
  formatQuantity: (symbol, quantity) => quantity.toFixed(8),
});

Environment variables (.env)

Before loading pine.module, the CLI loads .env files in the same order as for strategy modules — project root first, then the .pine file directory (overrides root):

my-project/
├── math/
│   ├── .env                          ← loaded second (overrides root)
│   └── impulse_trend_15m.pine
├── .env                              ← loaded first
└── package.json

Use this to store API keys without hardcoding them:

# .env
BYBIT_API_KEY=xxx
BYBIT_API_SECRET=yyy

// modules/pine.module.ts
addExchangeSchema({
  exchangeName: "my-exchange",
  getCandles: async (symbol, interval, since, limit) => {
    const exchange = new ccxt.bybit({
      apiKey: process.env.BYBIT_API_KEY,
      secret: process.env.BYBIT_API_SECRET,
      enableRateLimit: true,
    });
    // ...
  },
});

Then run:

npx @backtest-kit/cli --pine ./math/impulse_trend_15m.pine \
  --exchange my-exchange \
  --symbol BTCUSDT \
  --timeframe 15m \
  --limit 180 \
  --when "2025-09-24T12:00:00.000Z"

Or add it to package.json:

{
  "scripts": {
    "pine": "npx @backtest-kit/cli --pine ./math/impulse_trend_15m.pine --symbol BTCUSDT --timeframe 15m --limit 180"
  }
}

npm run pine

PineScript Requirements

The CLI reads all plot() calls that use display=display.data_window as output columns. Every other plot() is ignored. Name each output plot explicitly:

//@version=5
indicator("MyIndicator", overlay=true)

// ... computation ...

plot(close,    "Close",    display=display.data_window)
plot(position, "Position", display=display.data_window)

The column names in the output Markdown table are taken directly from those plot names — no manual schema definition needed.

Output

The CLI prints a Markdown table to stdout:

# PineScript Technical Analysis Dump

**Signal ID**: CLI execution 2025-09-24T12:00:00.000Z

| Close | Position | timestamp |
| --- | --- | --- |
| 112871.28 | -1.0000 | 2025-09-22T15:00:00.000Z |
| 112666.69 | -1.0000 | 2025-09-22T15:15:00.000Z |
| 112736.00 |  0.0000 | 2025-09-22T18:30:00.000Z |
| 112653.90 |  1.0000 | 2025-09-22T22:15:00.000Z |

Save to ./math/dump/impulse_trend_15m.md (uses .pine file name automatically, dump is created next to the .pine file):

npx @backtest-kit/cli --pine ./math/impulse_trend_15m.pine --markdown

Override the output name with --output:

npx @backtest-kit/cli --pine ./math/impulse_trend_15m.pine --jsonl --output feb2026_bb
# → ./math/dump/feb2026_bb.jsonl

Print to stdout (no flag):

npx @backtest-kit/cli --pine ./math/impulse_trend_15m.pine

🎨 Visual Pine Script Editor

@backtest-kit/cli ships a browser-based Pine Script editor powered by @backtest-kit/ui. It lets you write, run, and iterate on indicators interactively — with a live chart that updates as you hit ▶ Run

Usage

npx @backtest-kit/cli --editor

The CLI will:

1. Load ./modules/editor.module (if it exists) — use it to register your exchange schema, identical to pine.module
2. Start the @backtest-kit/ui server on http://localhost:60050 (or CC_WWWROOT_PORT)
3. Open http://localhost:{CC_WWWROOT_PORT}?pine=1 automatically in your default browser

Press Ctrl+C to stop the server.

Exchange via editor.module

Drop a modules/editor.module.ts next to your project to register the exchange that the editor's candle provider will use:

// modules/editor.module.ts
import { addExchangeSchema } from "backtest-kit";
import ccxt from "ccxt";

addExchangeSchema({
  exchangeName: "my-exchange",
  getCandles: async (symbol, interval, since, limit) => {
    const exchange = new ccxt.bybit({ enableRateLimit: true });
    const ohlcv = await exchange.fetchOHLCV(symbol, interval, since.getTime(), limit);
    return ohlcv.map(([timestamp, open, high, low, close, volume]) => ({
      timestamp, open, high, low, close, volume,
    }));
  },
  formatPrice: (symbol, price) => price.toFixed(2),
  formatQuantity: (symbol, quantity) => quantity.toFixed(8),
});

Environment Variables

| Variable | Default | Description | |-------------------|-----------|----------------------------------| | CC_WWWROOT_HOST | 0.0.0.0 | UI server bind address | | CC_WWWROOT_PORT | 60050 | UI server port |

package.json script

{
  "scripts": {
    "editor": "npx @backtest-kit/cli --editor"
  }
}

npm run editor

💾 Dumping Raw Candles

@backtest-kit/cli can fetch raw OHLCV candles from any registered exchange and save them to a file — no strategy file required.

CLI Flags

| Flag | Type | Description | |------|------|-------------| | --dump | boolean | Enable candle dump mode | | --symbol | string | Trading pair (default: "BTCUSDT") | | --timeframe | string | Candle interval (default: "15m") | | --limit | string | Number of candles to fetch (default: 250) | | --when | string | End date for candle window — ISO 8601 or Unix ms (default: now) | | --exchange | string | Exchange name (default: first registered, falls back to CCXT Binance) | | --output | string | Output file base name without extension (default: {SYMBOL}_{LIMIT}_{TIMEFRAME}_{TIMESTAMP}) | | --json | boolean | Write candles as a JSON array to ./dump/{output}.json | | --jsonl | boolean | Write candles as JSONL (one row per line) to ./dump/{output}.jsonl |

The dump/ directory is created in the current working directory (where the CLI is invoked from).

Exchange via dump.module

By default the CLI registers CCXT Binance automatically. To use a different exchange — or to configure API keys, custom rate limits, or a non-spot market — create a modules/dump.module.ts file. The CLI loads it automatically before fetching candles.

The CLI looks for modules/dump.module in the current working directory

my-project/
├── modules/
│   └── dump.module.ts            ← exchange registration
├── dump/                         ← auto-created: candle output files
└── package.json

Inside dump.module.ts call addExchangeSchema from backtest-kit:

// modules/dump.module.ts
import { addExchangeSchema } from "backtest-kit";
import ccxt from "ccxt";

addExchangeSchema({
  exchangeName: "my-exchange",
  getCandles: async (symbol, interval, since, limit) => {
    const exchange = new ccxt.bybit({ enableRateLimit: true });
    const ohlcv = await exchange.fetchOHLCV(symbol, interval, since.getTime(), limit);
    return ohlcv.map(([timestamp, open, high, low, close, volume]) => ({
      timestamp, open, high, low, close, volume,
    }));
  },
  formatPrice: (symbol, price) => price.toFixed(2),
  formatQuantity: (symbol, quantity) => quantity.toFixed(8),
});

Output

Each candle row contains OHLCV fields. Print to stdout:

npx @backtest-kit/cli --dump --symbol BTCUSDT --timeframe 15m --limit 100

Save to ./dump/BTCUSDT_100_15m_{timestamp}.jsonl:

npx @backtest-kit/cli --dump --symbol BTCUSDT --timeframe 15m --limit 100 --jsonl

Fetch candles up to a specific date with --when and override the file name with --output:

npx @backtest-kit/cli --dump --symbol BTCUSDT --timeframe 15m --limit 500 \
  --when "2026-02-28T00:00:00.000Z" \
  --jsonl --output feb2026_btc
# → ./dump/feb2026_btc.jsonl

Or add it to package.json:

{
  "scripts": {
    "dump": "npx @backtest-kit/cli --dump --symbol BTCUSDT --timeframe 15m --limit 500 --jsonl"
  }
}

npx @backtest-kit/cli --dump --symbol BTCUSDT --timeframe 15m --limit 500 --jsonl

🐞 PnL Debug (--pnldebug)

@backtest-kit/cli can simulate a hypothetical position minute by minute and print running PnL, peak profit, and maximum drawdown for each candle — without placing any trades or loading a strategy file.

CLI Flags

| Flag | Type | Description | |------|------|-------------| | --pnldebug | boolean | Enable PnL debug mode | | --priceopen | number | Entry price (required) | | --direction | string | long or short (default: long) | | --when | string | Start timestamp — ISO 8601 or Unix ms (default: now) | | --minutes | string | Number of 1m candles to simulate (default: 60) | | --symbol | string | Trading pair (default: "BTCUSDT") | | --exchange | string | Exchange name (default: first registered, falls back to CCXT Binance) | | --output | string | Output file base name (default: {SYMBOL}_{DIRECTION}_{PRICEOPEN}_{TIMESTAMP}) | | --json | boolean | Save results as JSON array to ./dump/<output>.json | | --jsonl | boolean | Save results as JSONL to ./dump/<output>.jsonl | | --markdown | boolean | Save results as Markdown table to ./dump/<output>.md |

Output columns

| Column | Description | |--------|-------------| | min | Minute offset from start (1-based) | | timestamp | Candle timestamp (ISO 8601) | | close | Candle close price | | pnl% | Running PnL vs entry price (signed %) | | peak% | Highest PnL reached so far (always ≥ 0) | | drawdown% | Lowest PnL reached so far (always ≤ 0) |

Exchange via pnldebug.module

By default the CLI registers CCXT Binance automatically. To use a different exchange, create a modules/pnldebug.module.ts file in the current working directory — the CLI loads it automatically before fetching candles.

// modules/pnldebug.module.ts
import { addExchangeSchema } from "backtest-kit";
import ccxt from "ccxt";

addExchangeSchema({
  exchangeName: "my-exchange",
  getCandles: async (symbol, interval, since, limit) => {
    const exchange = new ccxt.bybit({ enableRateLimit: true });
    const ohlcv = await exchange.fetchOHLCV(symbol, interval, since.getTime(), limit);
    return ohlcv.map(([timestamp, open, high, low, close, volume]) => ({
      timestamp, open, high, low, close, volume,
    }));
  },
  formatPrice: (symbol, price) => price.toFixed(2),
  formatQuantity: (symbol, quantity) => quantity.toFixed(8),
});

Usage

Print to stdout (default table format):

npx @backtest-kit/cli --pnldebug --symbol BTCUSDT --priceopen 64069.50 --direction short --when "2025-02-25" --minutes 120

Save as Markdown:

npx @backtest-kit/cli --pnldebug --priceopen 67956.73 --direction long --when 1772064000000 --minutes 60 --markdown
# → ./dump/BTCUSDT_long_67956.73_{timestamp}.md

Override the output file name with --output:

npx @backtest-kit/cli --pnldebug --priceopen 64069.50 --direction short --when "2025-02-25" --minutes 120 \
  --jsonl --output feb25_short_debug
# → ./dump/feb25_short_debug.jsonl

Or add it to package.json:

{
  "scripts": {
    "pnldebug": "npx @backtest-kit/cli --pnldebug --symbol BTCUSDT --priceopen 64069.50 --direction short --when \"2025-02-25\" --minutes 120"
  }
}

npm run pnldebug

Example stdout output

Symbol: BTCUSDT | Direction: short | PriceOpen: 64069.50 | From: 2025-02-25T00:00:00.000Z | Minutes: 120

  min | timestamp                 |        close  |    pnl%  |   peak%  | drawdown%
-----------------------------------------------------------------------------------
    1 | 2025-02-25T00:01:00.000Z  |      64020.10 |   +0.08% |   +0.08% |     0.00%
    2 | 2025-02-25T00:02:00.000Z  |      64105.30 |   -0.06% |   +0.08% |    -0.06%
  ...
  120 | 2025-02-25T02:00:00.000Z  |      63200.00 |   +1.36% |   +1.36% |    -0.06%

🐛 Broker Debug (--brokerdebug)

@backtest-kit/cli can fire a single broker commit against your live broker adapter without running a full strategy — useful for verifying that your brokerdebug.module correctly wires up exchange calls.

CLI Flags

| Flag | Type | Description | |------|------|-------------| | --brokerdebug | boolean | Enable broker debug mode | | --commit | string | Commit type to fire (default: "signal-open") | | --symbol | string | Trading pair (default: "BTCUSDT") | | --exchange | string | Exchange name (default: first registered) |

Available --commit values:

| Value | Broker hook | |-------|-------------| | signal-open | onSignalOpenCommit | | signal-close | onSignalCloseCommit | | partial-profit | onPartialProfitCommit | | partial-loss | onPartialLossCommit | | average-buy | onAverageBuyCommit | | trailing-stop | onTrailingStopCommit | | trailing-take | onTrailingTakeCommit | | breakeven | onBreakevenCommit |

How It Works

The CLI loads ./modules/brokerdebug.module, fetches the last candle for --symbol / --timeframe, derives synthetic payload values from currentPrice (TP = +2%, SL = -2%), and calls the selected broker hook once. Exits with code 0 on success.

Broker via brokerdebug.module

Create a modules/brokerdebug.module.ts file and register your broker adapter:

// modules/brokerdebug.module.ts
import { Broker } from 'backtest-kit';
import { myExchange } from './exchange.mjs';

class MyBroker {
  async onSignalOpenCommit({ symbol, priceOpen, position }) {
    await myExchange.openPosition(symbol, position, priceOpen);
  }
  // ... other hooks
}

Broker.useBrokerAdapter(MyBroker);
Broker.enable();

Usage

npx @backtest-kit/cli --brokerdebug --commit signal-open --symbol BTCUSDT
npx @backtest-kit/cli --brokerdebug --commit partial-profit --symbol ETHUSDT --timeframe 1h

🗑️ Flushing Strategy Output (--flush)

@backtest-kit/cli can delete generated output folders from one or more strategy dump directories without touching cached candle data.

CLI Flags

| Flag | Type | Description | |------|------|-------------| | --flush | boolean | Enable flush mode |

Positional arguments (required): one or more strategy entry point files. For each entry point the CLI resolves its directory and removes the following subdirectories from <entry-dir>/dump/:

| Folder | Contents | |--------|----------| | report | Backtest report files (.jsonl) | | log | Run logs (log.jsonl) | | markdown | Exported Markdown reports | | agent | Agent outline files |

Candle cache (dump/data/) and AI forecast outlines (dump/outline/) are not removed.

Usage

Flush a single strategy:

npx @backtest-kit/cli --flush ./content/feb_2026.strategy/modules/backtest.module.ts

Flush multiple strategies at once:

npx @backtest-kit/cli --flush \
  ./content/feb_2026.strategy/modules/backtest.module.ts \
  ./content/mar_2026.strategy/modules/backtest.module.ts

Or add it to package.json:

{
  "scripts": {
    "flush": "npx @backtest-kit/cli --flush ./content/feb_2026.strategy/modules/backtest.module.ts"
  }
}

npm run flush

🗂️ Scaffolding a New Project (--init)

@backtest-kit/cli can bootstrap a ready-to-use project directory with a pre-configured layout, example strategy files, and all documentation fetched automatically.

CLI Flags

| Flag | Type | Description | |------|------|-------------| | --init | boolean | Scaffold a new project | | --output | string | Target directory name (default: backtest-kit-project) |

Usage

npx @backtest-kit/cli --init

Creates ./backtest-kit-project/ in the current working directory.

Override the directory name with --output:

npx @backtest-kit/cli --init --output my-trading-bot

Creates ./my-trading-bot/.

The target directory must not exist or must be empty — the command aborts if it contains any files.

Generated Project Structure

backtest-kit-project/
├── package.json              # pre-configured with all backtest-kit dependencies
├── .gitignore
├── CLAUDE.md                 # AI-agent guide for writing strategies
├── content/
│   └── feb_2026.strategy.ts  # example strategy entry point
├── docs/
│   ├── lib/                  # fetched automatically (see below)
│   ├── backtest_actions.md
│   ├── backtest_graph_pattern.md
│   ├── backtest_logging_jsonl.md
│   ├── backtest_pinets_usage.md
│   ├── backtest_risk_async.md
│   ├── backtest_strategy_structure.md
│   ├── pine_debug.md
│   └── pine_indicator_warmup.md
├── math/
│   └── feb_2026.pine         # example PineScript indicator
├── modules/
│   ├── dump.module.ts        # exchange schema for --dump mode
│   └── pine.module.ts        # exchange schema for --pine mode
├── report/
│   └── feb_2026.md           # example strategy research report
└── scripts/
    └── fetch_docs.mjs        # utility: downloads library READMEs into docs/lib/

Automatic Documentation Fetch

After scaffolding, the CLI immediately runs scripts/fetch_docs.mjs inside the new project, which downloads the latest README files for all bundled libraries into docs/lib/:

| File | Source | |------|--------| | backtest-kit.md | backtest-kit README | | backtest-kit__graph.md | @backtest-kit/graph README | | backtest-kit__pinets.md | @backtest-kit/pinets README | | backtest-kit__cli.md | @backtest-kit/cli README | | garch.md | garch README | | volume-anomaly.md | volume-anomaly README | | agent-swarm-kit.md | agent-swarm-kit README | | functools-kit.md | functools-kit README |

You can re-run this script at any time to refresh the docs:

cd backtest-kit-project
node ./scripts/fetch_docs.mjs

Or via the pre-configured npm script:

npm run sync:lib

🐳 Running in Docker (--docker)

CLI can create a ready-to-use Docker workspace: self-contained directory with docker-compose.yaml and a strategy entry point.

CLI Flags

| Flag | Type | Description | |------|------|-------------| | --docker | boolean | Scaffold a Docker workspace | | --output | string | Target directory name (default: backtest-kit-docker) |

Usage

npx @backtest-kit/cli --docker

Creates ./backtest-kit-docker/ in the current working directory.

Override the directory name with --output:

npx @backtest-kit/cli --docker --output my-docker-workspace

The target directory must not exist or must be empty — the command aborts if it contains any files.

Two Launch Modes

The Docker image entrypoint supports two ways to run a strategy:

1. command: in docker-compose.yaml

Pin mode and flags directly in the compose file. The entrypoint forwards all arguments to the CLI unchanged:

command:
  - --live
  - --symbol
  - TRXUSDT
  - --strategy
  - feb_2026_strategy
  - --exchange
  - ccxt-exchange
  - ./content/feb_2026/feb_2026.strategy.ts
  - --ui

2. Inline environment variables

Pass MODE and STRATEGY_FILE on the command line — no file edits needed:

MODE=live SYMBOL=TRXUSDT STRATEGY_FILE=./content/feb_2026/feb_2026.strategy.ts docker-compose up -d

| Variable | Required | Default | Description | |----------|----------|---------|-------------| | MODE | yes | — | backtest | live | paper | walker | | STRATEGY_FILE | yes | — | Path to strategy entry point (relative to working_dir) | | SYMBOL | no | BTCUSDT | Trading pair | | STRATEGY | no | first registered | Strategy name | | EXCHANGE | no | first registered | Exchange name | | FRAME | no | first registered | Frame name (backtest only) | | UI | no | — | Any non-empty value enables --ui | | TELEGRAM | no | — | Any non-empty value enables --telegram | | VERBOSE | no | — | Any non-empty value enables --verbose | | NO_CACHE | no | — | Any non-empty value enables --noCache | | NO_FLUSH | no | — | Any non-empty value enables --noFlush | | ENTRY | no | — | Any non-empty value enables multiple symbols from userspace |

🌍 Environment Variables

Create a .env file in your project root:

# Telegram notifications (required for --telegram)
CC_TELEGRAM_TOKEN=your_bot_token_here
CC_TELEGRAM_CHANNEL=-100123456789

# Web UI server (optional, defaults shown)
CC_WWWROOT_HOST=0.0.0.0
CC_WWWROOT_PORT=60050

# Custom QuickChart service URL (optional)
CC_QUICKCHART_HOST=

| Variable | Default | Description | |------------------------|-------------|---------------------------------------| | CC_TELEGRAM_TOKEN | — | Telegram bot token (from @BotFather) | | CC_TELEGRAM_CHANNEL | — | Telegram channel or chat ID | | CC_WWWROOT_HOST | 0.0.0.0 | UI server bind address | | CC_WWWROOT_PORT | 60050 | UI server port | | CC_QUICKCHART_HOST | — | Self-hosted QuickChart instance URL |

⚙️ Default Behaviors

When your strategy module does not register an exchange, frame, or strategy name, the CLI falls back to built-in defaults and prints a console warning:

| Component | Default | Warning | |--------------|--------------------------------|---------------------------------------------------------------------------| | Exchange | CCXT Binance (default_exchange) | Warning: The default exchange schema is set to CCXT Binance... | | Frame | February 2024 (default_frame) | Warning: The default frame schema is set to February 2024... | | Symbol | BTCUSDT | — | | Cache intervals | 1m, 15m, 30m, 4h | Used if --cacheInterval not provided; skip entirely with --noCache |

Note: The default exchange schema does not support order book fetching in backtest mode. If your strategy calls getOrderBook() during backtest, you must register a custom exchange schema with your own snapshot storage.

🔧 Programmatic API

In addition to the CLI, @backtest-kit/cli can be used as a library — call run() directly from your own script without spawning a child process or parsing CLI flags.

run(mode, args)

import { run } from '@backtest-kit/cli';

await run(mode, args);

| Parameter | Description | |-----------|-------------| | mode | "backtest" \| "paper" \| "live" — Execution mode | | args | Mode-specific options (all optional — same defaults as CLI) |

run() can be called only once per process. A second call throws "Should be called only once".

Payload fields

Backtest (mode: "backtest"):

| Field | Type | Description | |-------|------|-------------| | entryPoint | string | Path to strategy entry point file | | symbol | string | Trading pair (default: "BTCUSDT") | | strategy | string | Strategy name (default: first registered) | | exchange | string | Exchange name (default: first registered) | | frame | string | Frame name (default: first registered) | | cacheInterval | CandleInterval[] | Intervals to pre-cache (default: ["1m","15m","30m","1h","4h"]) | | noCache | boolean | Skip candle cache warming (default: false) | | noFlush | boolean | Skip removing report/log/markdown/agent folders before the run (default: false) | | verbose | boolean | Log each candle fetch (default: false) |

Paper and Live (mode: "paper" / mode: "live"):

| Field | Type | Description | |-------|------|-------------| | entryPoint | string | Path to strategy entry point file | | symbol | string | Trading pair (default: "BTCUSDT") | | strategy | string | Strategy name (default: first registered) | | exchange | string | Exchange name (default: first registered) | | verbose | boolean | Log each candle fetch (default: false) |

Examples

Backtest:

import { run } from '@backtest-kit/cli';

await run('backtest', {
  entryPoint: './src/index.mjs',
  symbol: 'ETHUSDT',
  frame: 'feb-2024',
  cacheInterval: ['1m', '15m', '1h'],
  verbose: true,
});

Paper trading:

import { run } from '@backtest-kit/cli';

await run('paper', {
  entryPoint: './src