Skip to main content

Basic Usage

This guide helps you get started with OpenHuman.

Create Your First Agent

npx openhuman agent create --name "My Assistant"

Or using code:

import { OpenHuman } from 'openhuman-sdk';

const client = new OpenHuman({
apiKey: process.env.OPENHUMAN_API_KEY,
});

const agent = await client.agents.create({
name: 'My Assistant',
instructions: 'You are a helpful assistant who can answer a variety of questions.',
});

console.log('Agent ID:', agent.id);

Start a Conversation

const response = await client.agents.run(agent.id, {
message: 'Hello, please introduce yourself',
});

console.log(response.message);

Use Tools

Agents can use various tools to complete tasks:

const response = await client.agents.run(agent.id, {
message: 'Search for the latest AI news for me',
tools: ['web-search'], // Specify tools to use
});

Manage Memory

// Store information
await client.memory.store({
content: 'User prefers a minimalist design style',
source: 'conversation',
tags: ['preference', 'design'],
});

// Retrieve memories
const memories = await client.memory.recall({
query: "User's design preferences",
});

Configure an Agent

const agent = await client.agents.create({
name: 'Tech Documentation Assistant',
instructions: `
You are a professional technical documentation assistant.

Behavior guidelines:
- Answer concisely and clearly
- Use markdown formatting for code examples
- Be honest when you are uncertain about something
`,
tools: ['web-search', 'web-scraper'],
model: 'claude-sonnet-4-20250514',
});

Monitor Run Status

const run = await client.agents.run(agent.id, {
message: 'Write a blog post about AI',
});

console.log('Run ID:', run.id);
console.log('Status:', run.status);

// Get result
const result = await client.runs.get(run.id);
console.log(result.message);

Next Steps