This guide will help you get started with Truffle AI by creating and using your first AI agent.

​
Installation

Install the Truffle AI SDK using npm:

npm install truffle-ai

​
Basic Usage

​
1. Initialize the Client

Login to Truffle AI, go to settings page and get your API key.

Use this to create a new TruffleAI client:

import { TruffleAI } from 'truffle-ai';

const truffle = new TruffleAI('your-api-key');

​
2. Create an Agent

Deploy a new AI agent with specific instructions:

const agent = await truffle.deployAgent({
  name: 'Support Assistant',
  instruction: 'You are a customer support specialist who helps users with their questions.',
  model: 'gpt-4o-mini'
});

You can also create agents directly on the platform.

​
3. Using Your Agent

You can use your agent in two ways:

​
Run One-off Tasks

For single tasks or questions:

const result = await agent.run('What is artificial intelligence?');
console.log(result);

​
Chat Sessions

For interactive conversations with context:

// Create a chat session
const chat = agent.chat();

// Send messages and get responses
const response1 = await chat.send('Hi! My name is Alice.');
console.log(response1);

const response2 = await chat.send("What's my name?"); // Agent remembers "Alice"
console.log(response2);

// Get chat history
const history = chat.getHistory();

// Clear chat history if needed
chat.clearHistory();

​
Agent Management

​
Loading Existing Agents

Load a previously created agent:

const existingAgent = await truffle.loadAgent('your-agent-id');

You can see all your agents on the agents page. Get agent ids by clicking on the info button on the agent card.

​
Complete Example

Here’s a complete example showing various agent operations:

import { TruffleAI } from 'truffle-ai';

async function main() {
  // Initialize client
  const truffle = new TruffleAI('your-api-key');
  
  // Create an agent
  const agent = await truffle.deployAgent({
    name: 'Support Assistant',
    instruction: 'You are a customer support specialist who helps users with their questions.',
    model: 'gpt-4o-mini'
  });
  
  // Run a one-off task
  const taskResult = await agent.run('What are your main responsibilities?');
  console.log('Task Result:', taskResult);
  
  // Start a chat session
  const chat = agent.chat();
  
  // Have a conversation
  const responses = await Promise.all([
    chat.send('Hi! I need help with my account.'),
    chat.send('How do I change my password?'),
    chat.send('Thanks for your help!')
  ]);
  
  // Get chat history
  const history = chat.getHistory();
  console.log('Chat History:', history);
}

main().catch(console.error);

​
Next Steps

Now that you’ve created your first agent, you can:

Explore the SDK

Learn about all available features and options

View Examples

See practical examples and use cases

Agent Guide

Learn more about working with agents

This guide will help you get started with Truffle AI by creating and using your first AI agent.

​
Installation

Install the Truffle AI SDK using npm:

npm install truffle-ai

​
Basic Usage

​
1. Initialize the Client

Login to Truffle AI, go to settings page and get your API key.

Use this to create a new TruffleAI client:

import { TruffleAI } from 'truffle-ai';

const truffle = new TruffleAI('your-api-key');

​
2. Create an Agent

Deploy a new AI agent with specific instructions:

const agent = await truffle.deployAgent({
  name: 'Support Assistant',
  instruction: 'You are a customer support specialist who helps users with their questions.',
  model: 'gpt-4o-mini'
});

You can also create agents directly on the platform.

​
3. Using Your Agent

You can use your agent in two ways:

​
Run One-off Tasks

For single tasks or questions:

const result = await agent.run('What is artificial intelligence?');
console.log(result);

​
Chat Sessions

For interactive conversations with context:

// Create a chat session
const chat = agent.chat();

// Send messages and get responses
const response1 = await chat.send('Hi! My name is Alice.');
console.log(response1);

const response2 = await chat.send("What's my name?"); // Agent remembers "Alice"
console.log(response2);

// Get chat history
const history = chat.getHistory();

// Clear chat history if needed
chat.clearHistory();

​
Agent Management

​
Loading Existing Agents

Load a previously created agent:

const existingAgent = await truffle.loadAgent('your-agent-id');

You can see all your agents on the agents page. Get agent ids by clicking on the info button on the agent card.

​
Complete Example

Here’s a complete example showing various agent operations:

import { TruffleAI } from 'truffle-ai';

async function main() {
  // Initialize client
  const truffle = new TruffleAI('your-api-key');
  
  // Create an agent
  const agent = await truffle.deployAgent({
    name: 'Support Assistant',
    instruction: 'You are a customer support specialist who helps users with their questions.',
    model: 'gpt-4o-mini'
  });
  
  // Run a one-off task
  const taskResult = await agent.run('What are your main responsibilities?');
  console.log('Task Result:', taskResult);
  
  // Start a chat session
  const chat = agent.chat();
  
  // Have a conversation
  const responses = await Promise.all([
    chat.send('Hi! I need help with my account.'),
    chat.send('How do I change my password?'),
    chat.send('Thanks for your help!')
  ]);
  
  // Get chat history
  const history = chat.getHistory();
  console.log('Chat History:', history);
}

main().catch(console.error);

​
Next Steps

Now that you’ve created your first agent, you can:

Explore the SDK

Learn about all available features and options

View Examples

See practical examples and use cases

Agent Guide

Learn more about working with agents