Getting Started & Configuration

This page provides a comprehensive guide for setting up the news-sentiment-ai-trader codebase. It covers the installation of system requirements, environment configuration, and the initial execution of the trading system using the integrated CLI.

System Requirements

The system relies on a modern JavaScript runtime and specific AI/search providers.

Requirement	Version	Purpose
Node.js	≥15.0.0	Core runtime for backtest-kit and ccxt.
TypeScript	^5.0.0	Static typing and compilation.
Ollama	≥0.6.3	Local LLM server for sentiment inference.
Tavily API	-	Web search for real-time news retrieval.

---

Installation & Setup

1. Clone and Install Dependencies

The project uses npm to manage its dependencies, which include the backtest-kit framework and the agent-swarm-kit for LLM orchestration.

git clone https://github.com/tripolskypetr/backtest-kit.git backtest-kit-main
cd backtest-kit-main/example
npm install
Copy

The core logic depends on several specialized packages defined in the manifest:

• @backtest-kit/cli: Provides the command-line interface.
• agent-swarm-kit: Manages the multi-agent LLM forecasting logic.
• ollama: Interface for the local LLM.
• ccxt: Handles exchange connectivity for market data.

2. Environment Configuration

The system requires two primary API tokens to function. Copy the example environment file and populate it with your credentials:

cp .env.example .env
Copy

Required Variables:

• OLLAMA_TOKEN: Token for authenticating with your Ollama instance (if required by your setup).
• TAVILY_TOKEN: API key from Tavily for news search capabilities.

3. TypeScript Configuration

The project is configured to use ESNext modules with path mapping for the logic/ and utils/ directories to simplify imports.

---

Data Flow: Initialization to Execution

The following diagram illustrates the transition from environment configuration to the internal code entities that manage the system startup.

Configuration to Code Entity Mapping

"The diagram below bridges the environment variables and configuration files to the internal backtest-kit initialization sequence."

---

Running the System

CLI Execution

The system is started via the @backtest-kit/cli package, which is mapped to the start script in package.json.

npm start
Copy

Framework Initialization Sequence

When the system starts, it performs several critical steps via the backtest-kit framework:

1. Logger Setup: The framework requires a logger implementation to be set via setLogger to output operational data.
2. Global Configuration: Parameters such as CC_PERCENT_FEE and CC_SCHEDULE_AWAIT_MINUTES are applied to the GLOBAL_CONFIG.
3. Dependency Injection: Over 75 services are wired together using di-kit and di-scoped, managing the lifecycle of news fetchers, LLM clients, and backtesting engines.

Component Interaction Diagram

"This diagram maps the system's runtime components to their respective code modules during a typical execution cycle."

---

Verification

To ensure the environment is correctly configured, you can verify the framework initialization with a minimal script:

import { getDefaultConfig, setLogger } from 'backtest-kit';

// Verify Logger
setLogger({
  log: (topic, ...args) => console.log(topic, args),
  debug: () => {},
  info: () => {},
  warn: () => {},
});

// Verify Configuration
const config = getDefaultConfig();
console.log('Parameters loaded:', Object.keys(config).length);
Copy

Expected Result: The console should indicate that 14 default parameters have been loaded.