yahboom
HomeTutorialsProduct
Solution
Blog
Online Store
About Us
EN/中文
Store
Jetson Orin NANO Developer Kit📖
01.Get started
  • 01. Learning route
  • 02. Get started quickly
  • 03. System Update Notes
  • 04. Disabling automatic updates and troubleshooting issues
  • 05. Jetson Orin Nano official kit screwdriver instructions
02.Jetson basic course
  • 01.Jetson board introduction
  • 02.Write Jetson original system
  • 03.Installing Jetson environment
  • 04.Backup SSD system
  • 05.Restore initial image system
  • 06.Upgrade to SUPER version (Optional)
  • 07.Write SUPER boot (Must read)
  • 08.SSD expansion
  • 09.Jetpack old system upgraded to Yahboom SUPER factory system
  • 10.Write the system onto your own SSD
03.Basic setting course
  • 01.Basic knowledge
  • 02.Network configuration
  • 03.SSH remote login
  • 04.VNC remote control
  • 05.RDP remote control
  • 06.Remote file transfer
  • 07.Jtop tool
  • 08.Exchange space expansion
  • 09.Use VSCode
  • 10.Use VMware
  • 11.Modify screen resolution
04.GPIO control course
  • 01.GPIO description
  • 02.GPIO library installation
  • 03.GPIO reading
  • 04.GPIO output
  • 05.Serial communication
  • 06.I2C communication
05.Vision basic course
  • 01.CSI camera preview
  • 02.USB camera preview
  • 03.Use Jupyter Lab
  • 04.Use JetCam
06.OpenCV basic course
  • 01.Image reading
  • 02.Image saving
  • 03.Image modification
  • 04.Image scaling
  • 05.Image cropping
  • 06.Image translation
  • 07.Image mirroring
  • 08.Image grayscale conversion
  • 09.Image binarization
  • 10.Image edge detection
  • 11.Add line segments
  • 12.Add rectangles
  • 13.Add circular shape
  • 14.Add ellipses
  • 15.Add polygons
  • 16.Add Text
  • 17.Preview USB camera image
  • 18.Preview CSI camera image
07.Vision advanced course
01.Environment construction
  • 01.TensorFlow
  • 02.PyTorch
  • 03.Torchvision
02.DeepStream development
  • 01.DeepStream environment
  • 02.License plate detection and recognition
  • 03.Parallel multi-model inference
  • 04.Object embedding vector generation
  • 05.Pose recognition
  • 06.NV optical flow anomaly detection
  • 07.Single-view 3D tracking
  • 08.Multi-view 3D tracking
03.NVIDIA TAO toolkit application
  • 01.Traffic scene object detection
  • 02.Person object detection
  • 03.Vehicle brand classification
  • 04.Vehicle model classification
  • 05.License plate detection and recognition
  • 06.Person semantic segmentation
  • 07.Retail product detection and recognition
  • 08.Person feature extraction and matching (ReID)
  • 09.Person action and pose classification
  • 10.Transformer-based person detection
  • 11.3D human skeleton detection
  • 12.LiDAR 3D point cloud detection
04.YOLO11 development
  • 01.YOLO environment construction
  • 02.CLI use
  • 03.Object detection
  • 04.Instance segmentation
  • 05.Pose estimation
  • 06.Image classification
  • 07.Oriented object detection
  • 08.Model conversion
  • 09.Dataset annotation
  • 10.Model training conversion
  • 11.Model prediction
05.Mediapipe development
  • 01.Mediapipe environment
  • 02.Hand detection
  • 03.Face detection
  • 04.Face effects
  • 05.3D object recognition
  • 06.Virtual brush
  • 07.Gesture recognition
  • 08.Finger control
  • 09.Posture estimation
  • 10.Face mask
  • 11.Overall detection
  • 12.Green scene removal(Human)
06.YOLO26 development
  • 01.YOLO environment construction
  • 02.CLI use
  • 03.Object detection
  • 04.Instance segmentation
  • 05.Pose estimation
  • 06.Image classification
  • 07.Oriented object detection
  • 08.Model conversion
  • 09.Dataset annotation
  • 10.Model training conversion
  • 11.Model prediction
07.QR code recognition
  • QR code recognition
08.Docker basic course
  • 01.Docker introduction
  • 02.Docker installation
  • 03.Docker command
  • 04.Docker usage
  • 05.Docker interaction
09.ROS2-Jazzy
01.ROS2 basic course
  • 1.Introduction to ROS2
  • 2.ROS2 install Jazzy
  • 3.ROS2 development environment
  • 4.ROS2 workspace
  • 5.ROS2 function package
  • 6.ROS2 node
  • 7.ROS2 topic communication
  • 8.ROS2 service communication
  • 9.ROS2 action communication
  • 10.ROS2 custom interface message
  • 11.ROS2 parameter service case
  • 12.ROS2 meta-function package
  • 13.ROS2 distributed communication
  • 14.ROS2 DDS
  • 15.ROS2 time related API
  • 16.ROS2 common command tools
  • 17.ROS2 rviz2 use
  • 18.ROS2 rqt toolbox
  • 19.ROS2 Launch startup file configuration
  • 20.ROS2 recording and playback tool
  • 21.ROS2 URDF model
  • 22.ROS2 Gazebo simulation platform
  • 23.ROS2 TF2 coordinate transformation
02.ROS2 vision course
  • 01.Camera preview (USB)
  • 02.Camera preview (CSI)
  • 03.Camera calibration (USB)
  • 04.Camera calibration (CSI)
  • 05.AR vision (USB)
  • 06.AR vision (CSI)
10.Large model basic course
01.Offline large model
  • 01.AI large model environment deployment
  • 02.Install large model dialogue platform
  • 03.Meta AI - Llama model
  • 04.Alibaba Cloud - Qwen model
  • 05.Microsoft - Phi-4-mini model
  • 06.DeepSeek - DeepSeek model
  • 07.Alibaba Cloud - Qwen-VL visual multimodal large model
  • 08.Google - Gemma visual multimodal large model
  • 9.Llava visual multimodal large model
  • 10.MiniCPM-V visual multimodal large model
  • 11.High-quality vision-language grounding (LocateAnything)
02.Online large model
  • 01.Tongyi Qianwen large model
  • 02.OpenRouter large model API aggregation platform
11.Large model application practice
01.Voice interaction application
  • 1.Voice interaction hardware connection (Must read)
  • 2.Offline speech to text (ASR)
  • 3.Offline text to speech (TTS)
  • 4.AI large model voice interaction
  • 5.AI large model offline voice assistant
02.Multimodal large model application (Online_Offline large model + Text_Voice interaction)
  • 01.Multimodal visual understand application
  • 02.Multimodal text image application
  • 03.Multimodal video analysis application
  • 04.Multi module visual position application
  • 05.Multimodal table scanning application
  • 06.Multimodal autonomous agent application
12.NVIDIA Isaac ROS course
  • 01.Build Isaac ROS environment
  • 02.Deep segmentation
  • 03.DNN stereoscopic depth
  • 04.Free space segmentation
  • 05.Image distortion processing
  • 06.Image segmentation
  • 07.3D scene reconstruction and mapping
  • 08.Object detection
  • 09.3D attitude estimation
  • 10.Visual SLAM
13.OpenClaw deployment and basic usage
01.OpenClaw deployment and introduction
  • 01.Quick start with the pre-configured environment (Recommended)
  • 02.OpenClaw Deployment
  • 03.OpenClaw WhatsApp Plugin Application
  • 04.OpenClaw WebChat Interaction
  • 05.OpenClaw TUI Interaction
  • 06.OpenClaw Tools Introduction
  • 07.OpenClawGate Gateway User Manual
  • 08.OpenClaw Features Overview
  • 09.OpenClawHub Introduction (Skill Installation)
02.OpenClaw basic application
  • 01.OpenClaw Application - File Management
  • 02.OpenClaw Application - Camera
  • 03.OpenClaw Application - Browser Control
  • 04.OpenClaw Application - Script Execution
  • 05.OpenClaw Application - AI Programming (Peripheral & GPIO Control)
  • 06.OpenClaw Application - Dedicated AI Assistant
  • 07.OpenClaw Web Search Capability Configuration
03.NVIDIA official Agent Skills application
  • 01.AI-assisted development with VS Code or Cursor
  • 02.Jetson device skills
  • 03.Jetson BSP custom agent skills
14.OpenClaw advanced development
01.Preparation Before Use
  • 01.Peripheral Hardware Configuration
  • 02.OpenClaw API-KEY Configuration
  • 03.OpenClaw Model Switching
  • 04.OpenClaw Prompts
  • 05.AI Voice Interaction Configuration
  • 06.Three Dialogue Configuration Tests
02.OpenClaw AI Programming Peripheral Control
  • 1. Servo Control
  • 2. RGB Strip
  • 3. OLED Display
  • 4. Temperature and Humidity Sensor
  • 5. Camera Application
  • 6. AI Script Peripheral Control
03.OpenClaw Peripheral Integration Applications
  • 1. Plant Care Assistant
  • 2. AI Calorie Estimation
  • 3. AI Rock Paper Scissors Game
  • 4. AI Pet
  • 5. AI Weather Station
  • 6. Temperature Humidity Dashboard
  • 7. Scheduled Tasks
Download
  • System_File
  • VMFile_BootFile
  • 3D Model File
  • DataSheet
  • Software_Tools
  • Code
  • Manual
Download ZIP

Welcome to Jetson Orin NANO Developer Kit repository

03.Jetson BSP custom agent skills

Jetson BSP Customization Agent Skills Tutorial

1. Function Description

jetson-bsp-skills is a set of Jetson Linux BSP customization workflow Skills provided by NVIDIA. It targets host-side development before flashing the Jetson, and is not a device skill for daily diagnostics on a booted Jetson.

This repository does not contain BSP images, source code archives or specific carrier board materials. Instead, it provides workflow guidance, templates and reference materials to help developers prepare a BSP workspace, modify the BSP, build the related artifacts, and flash and verify after confirmation.

Official repository:

https://github.com/NVIDIA-AI-IOT/jetson-bsp-skills

2. Applicable Scenarios

When low-level customization of the Jetson platform is required, these Skills can be considered, for example:

  • Modifying Pinmux and GPIO pin multiplexing

  • Customizing USB, PCIe, UPHY, MGBE

  • Adding or modifying the device tree of a CSI camera

  • Adjusting clocks, DVFS, fan policies and nvpmodel

  • Adjusting the BSP reserved memory

  • Modifying the kernel, device tree or driver and rebuilding

  • Integrating the customization into the BSP image, then flashing and verifying

If you only install applications, deploy OpenClaw, run models or check the state of a booted system, BSP Skills are usually not required; device-side tools or jetson-device-skills can be used.

3. Working Location and Directory Relationships

BSP customization is performed in an independent workspace on a Linux host. The host can be a physical Ubuntu computer, or an Ubuntu virtual machine configured with sufficient disk space and resources.

Example directory:

xxxxxxxxxx
~/jetson-bsp-workspace/
├── .claude/              # BSP Skills, context and reference materials
├── Linux_for_Tegra/      # Prepared BSP flashing directory, determined by the selected workflow
└── source/               # Kernel, device tree and other source code initialized when needed

Claude Code should be started from the BSP workspace, not from a directory of the running system on the Jetson. The Linux_for_Tegra here is the to-be-flashed BSP working directory on the host, not the root file system running on the Jetson.

4. Host Environment Preparation

Prepare the following before starting:

  • An Ubuntu Linux host or an Ubuntu virtual machine

  • An available network connection and sufficient disk space

  • BSP materials matching the target Jetson model, carrier board and JetPack/L4T version

  • Claude Code and an available Anthropic login/authorization

  • If actual flashing is planned: a matching power supply, a USB data cable and a Recovery USB device that can be passed through directly

  • A backup of important data and a recoverable original system plan

The virtual machine can be used for BSP preparation and source code work. Before actual flashing, you must also confirm that the virtual machine can recognize and stably pass through the Jetson Recovery USB; if the USB device cannot be passed through reliably, do not start flashing.

5. Obtain the BSP Skills Repository

Execute the following in the Ubuntu host terminal:

xxxxxxxxxx
git clone https://github.com/NVIDIA-AI-IOT/jetson-bsp-skills.git
cd jetson-bsp-skills

If the repository already exists, enter it and check first instead of cloning it again:

xxxxxxxxxx
cd ~/jetson-bsp-skills
git status
git remote -v

6. Create the BSP Workspace and Install the Skills

Create an independent directory for this BSP work:

xxxxxxxxxx
mkdir -p ~/jetson-bsp-workspace

Run the installation script provided by the repository to copy the Skills, context and reference materials into the workspace:

xxxxxxxxxx
cd ~/jetson-bsp-skills
./setup.sh --workspace ~/jetson-bsp-workspace

The installation script creates the following in the workspace:

xxxxxxxxxx
~/jetson-bsp-workspace/.claude/

If this directory already exists, the script refuses to overwrite it by default. Confirm its content first; consider --force only when you clearly want to rebuild and have made a backup. Do not add --force directly just to eliminate an error, because the script deletes the existing .claude/ and recreates it.

Check the installation result:

xxxxxxxxxx
find ~/jetson-bsp-workspace/.claude -maxdepth 3 -type f | sort | head -80

7. Start Claude Code in the Workspace

xxxxxxxxxx
cd ~/jetson-bsp-workspace
claude

On the first start, complete the login or authorization as prompted by Claude Code. Enter the official recommended entry request in Claude Code:

xxxxxxxxxx
help me setup the bsp customization workspace

This request leads to the /jetson-quick-start workflow. Follow the prompts to choose automatic preparation or guided preparation based on existing materials, and provide the real Jetson model, carrier board, BSP/JetPack version and customization goal.

Do not just tell the agent "help me DIY a BSP". First explain the exact model of the target board, the carrier board model, the current/target JetPack version, the storage medium and the specific hardware requirements.

8. BSP Workflow Stages

The repository divides the work into the following stages:

8.1 Preparation

Create or select the target configuration according to the device and BSP version, prepare the BSP input, unpack the image, initialize the source code/toolchain, and register local reference materials. Common entries include:

  • /jetson-quick-start

  • /jetson-init-target

  • /jetson-set-target

  • /jetson-download-bsp

  • /jetson-init-image

  • /jetson-init-source

  • /jetson-link-docs

  • /jetson-generate-kb

After preparation is complete, whether Linux_for_Tegra, the source directory and the knowledge base are generated in the workspace depends on the selected workflow and the provided BSP input; do not assume that starting Claude Code automatically generates a complete flashing image.

8.2 Customization

Invoke the corresponding /jetson-customize-* Skill according to the requirement, for example pinmux, USB, PCIe, UPHY, camera, clock, fan or power mode. When using a custom carrier board, first confirm whether /jetson-derive-carrier is required.

Before each modification, confirm:

  • The specific board model and hardware version corresponding to the modification

  • The paths of the original files and the target files

  • That the configuration/device tree matches the target BSP version

  • The change content and potential impact

  • Whether an original backup and a rollback method are available

8.3 Build and Integration

If the modification involves kernel-side source code, use /jetson-build-source to build the DTB, modules or kernel artifacts. Then use /jetson-promote-image to integrate the customized files and build artifacts into the to-be-flashed BSP image.

If the change only affects the rootfs or the configuration, the build step may be different; follow the workflow of the corresponding Skill and the current BSP version.

8.4 Flashing and Verification

Only after static checks, difference checks and target confirmation are completed should you consider:

  • /jetson-flash-image: Flash the prepared BSP to the Jetson

  • /jetson-validate-image: Post-flash static check and device-side verification

Flashing may erase the target storage, overwrite the QSPI/boot components, or make the device temporarily unable to boot. Before starting, you must clearly confirm the target Jetson, the target storage device, the Recovery state, the BSP version and the backup; you cannot infer that flashing will not affect other storage only from "an empty SSD". The flashing medium and whether QSPI is involved are determined by the target board model and the corresponding BSP flashing solution.

9. Suggestions for the First Experience

On the first use, do a workflow experience that does not write to the disk first:

  1. Install the Skills in the Ubuntu workspace.

  2. Start Claude Code and run help me setup the bsp customization workspace.

  3. Provide accurate board model and version information.

  4. Let the agent explain the Setup workflow it intends to choose and the inputs required.

  5. Use /jetson-print-bsp-info to check the existing Linux_for_Tegra directory (if it has already been prepared).

  6. View the customization plan and the file differences first, without performing flashing.

  7. Only after becoming familiar with the workflow and preparing a recoverable environment should you evaluate building and deployment separately.

If you only want to experience the workflow, do not connect the Recovery device, do not run /jetson-flash-image, and do not let the agent execute commands that erase or overwrite storage.

10. Common Problems

10.1 setup.sh Reports That --workspace Is Required

The target workspace must be specified, for example:

xxxxxxxxxx
./setup.sh --workspace ~/jetson-bsp-workspace

10.2 It Reports .claude/ already exists

The script protects the existing directory by default and does not overwrite it. First check whether it already contains work content; only use --force to rebuild after confirming that it can be discarded and backed up.

10.3 Claude Code Cannot Find the Skills

Confirm that the current directory is the workspace specified during installation, and start Claude Code from that directory:

xxxxxxxxxx
cd ~/jetson-bsp-workspace
claude

Check whether .claude/skills and the SKILL.md of each Skill exist in the workspace.

10.4 The Virtual Machine Cannot See the Recovery Device

First check the USB passthrough settings and the device connection state of the virtual machine. Plan the flashing only after confirming that the Recovery USB device is visible in Ubuntu. When the device does not appear stably in the virtual machine, do not run the flashing workflow.

11. Risk Notice

  • BSP Skills provide workflow guidance and cannot replace NVIDIA official documentation, board schematics and engineering review.

  • All generated plans, commands, file differences and submitted content require manual review.

  • Choosing the wrong board model, carrier board, BSP version or target disk may make the Jetson unable to boot or overwrite the wrong device.

  • Errors in the device tree, kernel, Pinmux and power configuration may cause peripheral failure, boot failure or abnormal hardware behavior.

  • Flashing and overwriting are high-risk operations; execute them only after confirming the target, the backup and the recovery path.

  • Verify the compatibility of the current Skill and BSP version first, and do not directly apply the flashing command from an old tutorial to other versions.

Jetson device skills
Peripheral Hardware Configuration

COMPANY

  • About us
  • Team Style
  • Blog

LEARNING & EXPLORATION

  • Tutorials
  • Edge AI hardware
  • Educational robots
  • Embedded intelligent hardware

WHERE TO BUY

  • Yahboom Online Store
  • Amazon
  • Alibaba
  • AliExpress
  • Mercado

TECHNOLOGY SUPPORT

  • Email: support@yahboom.com

BE A DISTRIBUTOR

  • Email: sales@yahboom.com
QR Code

Copyright © 2017-2026 Yahboom. All Rights Reserved.

EN/中文