Getting started with your Heyron agent

Your first hour: open your agent, say hello, give it a first useful job, and connect the one app you use most.

In this tutorial

Your Heyron agent is an AI assistant with its own computer in the cloud. It has its own files, memory and terminal, and it keeps them between conversations. You don't need to set up a server or paste in any API keys. Your first hour is about getting to know it and giving it one useful job.

Open your agent

  1. Sign in at heyron.ai and you'll land on Agents.
  2. If your workspace is empty, you'll see a Ready to launch card. Click Launch my agent. It takes a moment to start.
  3. Click your agent's name, or Open, to go to its chat.

New agents start as Untitled agent. Click the pencil next to the name to give it one. You can also rename it later from the agent's Settings.

Say hello

The chat opens with What can I help with? Type in the Ask anything... box like you'd message a person. A good first message tells the agent who you are and what you're working on:

Your agent saves what it learns about you to memory files, so you don't have to repeat yourself next time. More on that in how your agent's memory works.

Give it one small, real job

The fastest way to see what your agent can do is a task with a clear finish line. Not "run my business", but something you can check in five minutes:

  • "Research the top five competitors for an invoicing app for freelancers. Save a comparison table to a file and link your sources."
  • "Draft three versions of a welcome email for new customers. Keep each under 120 words."
  • "Read this PDF and give me the five decisions I need to make." (Drag the file into the chat.)

Then check the result. If it's off, say what's wrong and ask for a fix. Small, specific jobs build trust much faster than big vague ones. How to talk to your agent has more on this.

What your agent can and can't do

Out of the box
  • Research the web and read pages
  • Write, edit and summarise
  • Write and run code on its own computer
  • Create and organise files
  • Remember your preferences between chats
  • Run jobs on a schedule when you ask
Needs you to connect something first
  • Read or send your email (connect Gmail)
  • See your calendar or documents
  • Message you on Telegram or WhatsApp
  • Work in your GitHub repositories
  • Act in any other app you use
  • Anything that needs your login

Your agent can't see your own computer's files. To share something, drag it into the chat or upload it in Files.

Connect the one app you use most

Pick one connection that makes your agent useful day to day:

One connection you actually use beats ten you set up and forget.

Your first week

  • Day 1: say hello, give it one job, connect one app.
  • Day 2–3: give it a personality so it sounds like your assistant, not a generic chatbot. See give your agent a personality.
  • Day 4–7: hand it one repeating job, like a morning summary or a weekly report. See scheduled tasks.

If you feel overwhelmed, shrink the job. One finished small task is worth more than a big plan.

Frequently asked questions

Did I create a second agent by accident?

Check Agents. Each row is one agent, with its own memory and files. A new chat inside the same agent is not a new agent: it's the same assistant, starting a fresh conversation.

Why does my agent say something different on Telegram than in the dashboard?

Each place you talk to it, like the dashboard or a Telegram chat, is its own conversation. It's the same agent with the same memory files, but it doesn't share the chat history between them.

Do I need my own API key?

No. Your membership includes $15 of AI usage each month, and you can add a top-up if you need more. See AI usage, models and top-ups.

What's the Dashboard button at the top of my agent?

It opens your agent's own control panel (OpenClaw's or Hermes') in a new tab. You won't need it for everyday use; it's there for advanced settings.

Still stuck?

Open a ticket and a person will help. Tell us what you tried and what you saw.

Contact support