Skip to content
intellibrix

Getting started

intellibrix requires Node 22 or newer.

Install

Section titled “Install”
Terminal window
npm install intellibrix @intellibrix/ai-sdk zod @ai-sdk/anthropic

@intellibrix/ai-sdk connects intellibrix to model vendors through the AI SDK. The vendor package (@ai-sdk/anthropic, @ai-sdk/openai, @ai-sdk/google, …) is an optional peer; install the one you use. To scaffold a whole project instead, run npx @intellibrix/cli init my-agent.

Your first brick

Section titled “Your first brick”
import { provider } from '@intellibrix/ai-sdk'
import { Brick, Intelligence, defineTool } from 'intellibrix'
import { z } from 'zod'


const brick = new Brick({
  name: 'assistant',
  intelligence: new Intelligence({
    provider: provider('anthropic:claude-sonnet-5'),
    model: 'claude-sonnet-5',
    system: 'You are a concise assistant.',
  }),
})


brick.use([
  defineTool({
    name: 'get_time',
    description: 'Get the current time as an ISO string',
    input: z.object({}),
    execute: async () => new Date().toISOString(),
  }),
])


const { text } = await brick.ask('What time is it?')
console.log(text)

Set ANTHROPIC_API_KEY and run it with npx tsx first.ts. The model sees the get_time tool, calls it, and answers from the result. The loop that does this (call the model, run the tools it asks for, feed results back) is runAgentLoop; you never write it yourself.

Run it without an API key

Section titled “Run it without an API key”

Every example in these docs can be exercised offline with a scripted model. This is also how you will test your own agents (see Testing agents):

import { Brick, Intelligence } from 'intellibrix'
import { MockProvider } from 'intellibrix/testing'


const brick = new Brick({
  name: 'assistant',
  intelligence: new Intelligence({
    provider: new MockProvider([{ text: 'Hello from a mock model.' }]),
    model: 'mock',
  }),
})


console.log((await brick.ask('Hi')).text)

Next steps

Section titled “Next steps”