Tzafon

Northstar CUA Fast is Tzafon's vision language model trained with reinforcement learning for computer use tasks. It implements a CUA (computer use agent) loop that predicts browser actions from screenshots, enabling AI agents to interact with web interfaces.

The model is accessed via Tzafon's Lightcone API platform. By integrating Tzafon with Kernel, you can run these AI-powered browser automations on cloud-hosted infrastructure using Kernel's Computer Controls API, eliminating the need for local browser management and enabling scalable, reliable AI agents.

Quick setup with our Tzafon example app

Get started quickly with our Kernel app template that includes a pre-configured Tzafon Northstar CUA Fast integration:

kernel create --name my-tzafon-app --template tzafon

Choose TypeScript or Python as the programming language.

Then follow the deploy and invoke guides to deploy and run your Tzafon automation on Kernel's infrastructure.

Build your own agent

For full control over the loop, drive Northstar CUA Fast from TypeScript with @onkernel/cua-agent:

import Kernel from "@onkernel/sdk";
import { CuaAgent } from "@onkernel/cua-agent";

const client = new Kernel({ apiKey: process.env.KERNEL_API_KEY! });
const browser = await client.browsers.create({ stealth: true });

const agent = new CuaAgent({
  browser,
  client,
  initialState: {
    model: "tzafon:tzafon.northstar-cua-fast",
    systemPrompt: "You are a careful browser automation agent.",
  },
});

await agent.prompt("Open news.ycombinator.com and summarize the top story.");

Benefits of using Kernel with Tzafon Northstar CUA Fast

• No local browser management: Run Northstar CUA Fast automations without installing or maintaining browsers locally
• Scalability: Launch multiple browser sessions in parallel for concurrent AI agents
• Stealth mode: Built-in anti-detection features for reliable web interactions
• Session state: Maintain browser state across runs via Profiles
• Live view: Debug your Tzafon agents with real-time browser viewing
• Cloud infrastructure: Run computationally intensive AI agents without local resource constraints

Next steps

• Check out live view for debugging your automations
• Learn about stealth mode for avoiding detection
• Learn how to properly terminate browser sessions
• Learn how to deploy your Tzafon app to Kernel
• Read the Lightcone API documentation for model details