---
title: "Watch: An AI Agent Calls a Live API From Its Sandbox"
slug: "agent-calls-live-api-from-sandbox"
date: "2026-11-04"
tag: "Application"
summary: "A three-minute video of an ibl.ai agent doing real work from inside its Linux sandbox: it reaches a live external API through a host the organization allowlisted, using a secret the agent can call but never read, and brings the result back into the conversation."
author: "ibl.ai Engineering"
repo: "iblai/vibe"
linkedin: |
  Here is what an AI agent with a real machine and a locked door looks like, in three minutes.

  The agent in this video runs inside its own Linux sandbox on ibl.ai. By default that machine has no network at all. The organization opens one host, binds one secret, and the agent calls a live API from inside the sandbox, uses the key without ever being able to read it, and brings the answer back into the chat.

  That is the shape we think agentic AI has to take in a university, a hospital or an agency: the agent can act, the organization decides exactly what it can reach, and every call is logged. No open internet, no keys pasted into prompts, no vendor in the loop.

  With ibl.ai you own all the code and the data, the sandbox included.

  #iblai #AgenticAI #EnterpriseAI #AIAgents #Sandbox
---

**An ibl.ai agent can now be watched doing real work from inside its sandbox: in this three-minute video it calls a live external API through a host the organization allowlisted, with a secret it can use but never read, and on ibl.ai you own all the code and the data, the sandbox included.**

<div style="max-width:720px; margin:2rem auto;">
  <div style="position:relative; width:100%; aspect-ratio:16/9; border-radius:16px; overflow:hidden; background:#000; box-shadow:0 8px 30px rgba(0,0,0,0.12);">
    <iframe src="https://www.youtube.com/embed/ioXZP-nljKw" title="ibl.ai: AI agent calls a live API from a sandbox" style="position:absolute; top:0; left:0; width:100%; height:100%; border:0;" allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture" allowfullscreen loading="lazy"></iframe>
  </div>
  <p style="text-align:center; font-size:0.85rem; color:#5f6368; margin-top:0.75rem;">ibl.ai: AI agent calls a live API from a sandbox (2:58). Also on <a href="https://www.youtube.com/watch?v=ioXZP-nljKw">YouTube</a>.</p>
</div>

## What does the video show?

The agent runs in the sandbox type that shipped in September: a real Linux virtual machine bound to the agent, with no network by default. The video walks through the three controls that let it reach the outside world safely, then shows the call happening.

- **A network policy** names the one host the agent may reach. Everything else stays closed.
- **A secret** holds the API key. The agent can use it in a request; it cannot print it, read it or send it anywhere else.
- **The call itself** runs from inside the sandbox, and the result comes back into the conversation like any other tool output.

## Why does it matter?

An agent with a real machine is useful. An agent with a real machine and open internet access is a risk no regulated organization will accept. The sandbox gives you the first without the second: the organization decides, host by host, what its agents can touch, and every call is logged.

The controls in the video are the ones described in [Agent Sandboxes: A Real Linux VM, Locked to Hosts You Allow](/updates/agent-sandbox-virtual-machines-network-policies) and [Agent Network Policies and Secrets, Now in Settings](/updates/virtual-machine-settings-policies-and-secrets). They are part of [Agentic OS](/product/agentic-os), and like the rest of the platform they run inside your own perimeter: on ibl.ai you own all the code and the data.