Build Your Agent Platform

Let's build an agent platform for your team. You control how agents are built, where they run, where their data lives, and how they use that data to improve.

From zero to production

1. Set up your platform: start AgentOS and connect the UI.
2. Build your first agent: build an agent that compiles a daily newletter for you.
3. Deploy to Render: publish your platform, secure it, and run the daily brief on a schedule.
4. How to use your agents: connect through the UI, API, or MCP.
5. Improve your agents: use simulations, traces, and evaluations to improve your agents.
6. Add coding agents: enable coding agents that can publish PRs; add sandboxes and checkpoints.
7. Company OS: automate company operations by turning agents into product owners.
8. Connect Slack: make your agents available on Slack.
9. Scale your platform: scale to a multi-replica setup and use redis for coordination.
10. Your platform is ready: check the setup, choose your next workflow, and get help.

Before you start

You'll need Docker, an OpenAI API key, and a coding agent.

Set up your agent platform

Give your coding agent this prompt:

Help me set up my agent platform and build my first agent.

Clone https://github.com/agno-agi/agentos-render into a folder called agent-platform, cd in, and run the setup-platform skill (in .agents/skills/).

The skill configures your local environment, starts AgentOS and PostgreSQL, and checks the API and MCP connection. Add your model credentials to your local environment file when prompted.

Connect the UI

Open os.agno.com, sign in, and choose Connect OS. Use the following connection details:

If the connection fails, ask your coding agent to check the running containers and API health.

After you connect, you should see the AgentOS UI where you can chat with your agents, inspect sessions and traces, and manage your agent activity from one place.

Next

Build your first agent: a short newspaper about the companies, products, and topics you care about.