Skip to content
Python JavaScript

Quick Start Tutorial (JavaScript/TypeScript)

Welcome to GraphBit for JavaScript! This tutorial will guide you through creating your first AI agent in just 5 minutes.

Prerequisites

Before starting, ensure you have: - Node.js 16+ installed - GraphBit installed (npm install @infinitibit_gmbh/graphbit) - An OpenAI API key set in your environment

Your First Agent

Let's create a simple agent that can answer questions.

Step 1: Basic Setup

Create a new TypeScript file quickstart.ts:

import 'dotenv/config';
import { init, LlmConfig, AgentBuilder } from '@infinitibit_gmbh/graphbit';

// Initialize the library
init();

// Configure LLM (using OpenAI)
const config = LlmConfig.openai({
  apiKey: process.env.OPENAI_API_KEY || '',
  model: 'gpt-4o-mini',
});

Step 2: Create and Build the Agent

// Create an agent builder
const builder = new AgentBuilder('Quickstart Agent', config)
  .description('A helpful assistant')
  .systemPrompt('You are a helpful AI assistant. Answer concisely.')
  .temperature(0.7);

// Build the agent
// Note: build() is async
const agent = await builder.build();

Step 3: Execute the Agent

async function main() {
  const input = 'What is the capital of France?';
  console.log(`User: ${input}`);

  try {
    // Execute the agent
    const response = await agent.execute(input);
    console.log(`Agent: ${response}`);
  } catch (error) {
    console.error('Execution failed:', error);
  }
}

main().catch(console.error);

Complete Example

Here's the complete working example:

import { init, LlmConfig, AgentBuilder } from '@infinitibit_gmbh/graphbit';

async function main() {
  // Initialize
  init();

  // Configure LLM
  const apiKey = process.env.OPENAI_API_KEY;
  if (!apiKey) {
    console.error('Please set OPENAI_API_KEY environment variable');
    return;
  }

  const config = LlmConfig.openai({
    apiKey,
    model: 'gpt-4o-mini',
  });

  // Build Agent
  console.log('Building agent...');
  const agent = await new AgentBuilder('Quickstart Agent', config)
    .description('A helpful assistant')
    .systemPrompt('You are a helpful AI assistant.')
    .build();

  // Execute
  const input = 'Explain quantum computing in one sentence.';
  console.log(`\nUser: ${input}`);

  const response = await agent.execute(input);
  console.log(`Agent: ${response}`);
}

main().catch(console.error);

Run Your Agent

export OPENAI_API_KEY="your-api-key-here"
npx ts-node quickstart.ts

Expected output: 

Building agent...

User: Explain quantum computing in one sentence.
Agent: Quantum computing uses the principles of quantum mechanics to process information in ways that classical computers cannot, enabling faster solutions to complex problems.

Working with Different LLM Providers

GraphBit supports multiple LLM providers. You can easily switch between them:

// Anthropic
const anthropicConfig = LlmConfig.anthropic({
  apiKey: process.env.ANTHROPIC_API_KEY || '',
  model: 'claude-3-5-sonnet-20241022',
});

// Ollama (Local)
const ollamaConfig = LlmConfig.ollama({
  model: 'llama2',
  baseUrl: 'http://localhost:11434',
});

// Use the config when building your agent
const agent = await new AgentBuilder('Local Agent', ollamaConfig).build();

Next Steps