NVIDIA-AI-IOTNVIDIA-AI-IOTOfficialยท2 Agent Skills Included

DeepStream_Coding_Agent

DeepStream video analytics pipelines and vision model benchmarking

Generates NVIDIA DeepStream video analytics pipelines with correct GStreamer plugins, TensorRT inference, tracking, and Kafka messaging. Imports object detection models from HuggingFace or NGC, builds TensorRT engines, and runs multi-stream benchmarks automatically. Eliminates manual config debugging and produces ready-to-review code plus PDF benchmark reports.
npx skills add NVIDIA-AI-IOT/DeepStream_Coding_Agent --all -g -y

All Skills in This Repository (2)

Pure Emerald Level Indicators

Frequently Asked Questions

FAQPage Schema
How to install DeepStream Coding Agent?โ–ผ

Run `npx skills add NVIDIA-AI-IOT/DeepStream_Coding_Agent --all -g -y` in your terminal to install both DeepStream skills globally.

How to build a DeepStream pipeline with AI?โ–ผ

The deepstream-dev skill gives your coding assistant curated DeepStream SDK 9.0 references, so it generates correct pyservicemaker pipelines with proper plugins, configs, and error handling.

How to import a HuggingFace model into DeepStream?โ–ผ

The deepstream-import-vision-model skill downloads the ONNX model, builds a TensorRT engine, creates a custom bbox parser, and runs validation and multi-stream benchmarks fully unattended.

Does it work with Claude Code and Cursor?โ–ผ

Yes. Both skills follow the standard SKILL.md format and activate automatically in Claude Code, Cursor, Codex, and other compatible coding assistants.

Do I need a GPU to generate DeepStream code?โ–ผ

No. Code generation works on any machine, but running the generated pipelines or model benchmarks requires an NVIDIA GPU with DeepStream SDK 9.0, CUDA 13.1, and TensorRT.

Related Repositories in Software Engineering

View All in Software Engineeringโ†’