How to Set Up NVIDIA Jetson for Remote AI Development with Codex

How do you prepare an NVIDIA Jetson for Agent Skills using Codex? This tutorial shows how to build a secure, remote Jetson development environment with SSH, Wi-Fi, Docker, and CUDA. In this video, you'll learn how to: - Configure secure remote SSH access over USB and setup Wi-Fi - Install, authenticate, and register Codex as a remote project - Add NVIDIA Jetson device skills for device management - Inspect existing driver and Docker configurations - Configure non-root Docker permissions and validate GPU access with a CUDA container test By the end of this guide, you’ll have a fully connected Codex project and a validated NVIDIA Jetson environment ready for AI-assisted development. Part 2: https://youtu.be/r_VP4E_Y3as ➡️ Learn more: https://www.jetson-ai-lab.com/tutorials/ai-assisted-development-on-jetson/ 📝 Tech blog: https://developer.nvidia.com/blog/nvidia-jetpack-7-2-1-adds-agentic-video-skills-and-t3000-emulation/ 00:00 – Introduction and hardware setup 00:17 – Set up SSH over USB 01:05 – Connect Jetson to Wi-Fi 01:48 – Install Codex CLI and prepare a project 02:28 – Add Jetson to the Codex desktop app 03:06 – Inspect the system and GPU readiness 03:16 – Install NVIDIA Jetson Device Skills 03:31 – Prepare GPU containers without the full JetPack SDK 03:57 – Configure Docker permissions 04:21 – Verify CUDA GPU access in a container 04:41 – Wrap up and next steps