Developer Machine Setup
Set up your development machine with the Wendy CLI
CLI Installation
Install the Wendy CLI for your development machine.
curl -fsSL https://install.wendy.dev/cli.sh | bashVerify the installation:
wendy --versioncurl -fsSL https://install.wendy.dev/cli.sh | bashVerify the installation:
wendy --versionInstall via winget:
winget install WendyLabs.Wendy --source wingetTo upgrade to the latest version:
winget upgrade WendyLabs.Wendy --source wingetVerify the installation:
wendy --versionThe wendy CLI is your developer tool for building, discovering, and deploying apps to Wendy devices. It helps set up WendyOS on NVIDIA Jetson and Raspberry Pi, and can target Headless Mac beta on Apple Silicon Macs.
AI Coding Agent Setup
Install the Wendy plugin if you want Codex or Claude Code to create and run Wendy apps for you. The Git source is wendylabsinc/wendy-agentic-coding; in Codex, the plugin appears as @Wendy.
codex plugin marketplace add wendylabsinc/wendy-agentic-codingThen open Codex and install it from /plugins.
claude plugin marketplace add wendylabsinc/wendy-agentic-coding && claude plugin install wendy-agentic-coding@wendy-agentic-codingVS Code Extension
Prefer to work from your editor? The Wendy for VS Code extension brings device discovery, one-click deployment, integrated Swift (LLDB) and Python (debugpy) debugging, agent updates, and WendyOS flashing directly into your IDE. It also works in VS Code forks like Cursor and Windsurf via the Open VSX Registry.
Install it from the Visual Studio Code Marketplace, or from the command line:
code --install-extension wendylabsinc.wendy-vscodeSee the VSCode Extension guide for setup and full feature details.
Next Steps
Choose your first deployment target. Dedicated edge hardware runs WendyOS; existing Linux and Apple Silicon machines run Wendy Agent.
Install WendyOS on edge hardware
Jetson Orin Nano or AGX Orin
Build GPU-accelerated edge AI applications with NVIDIA CUDA support.
Jetson AGX Thor
Target NVIDIA's highest-performance platform for physical AI workloads.
Raspberry Pi 3, 4, or 5
Use a compact ARM64 target for development, automation, or robotics.
ESP32
Build and deploy regular native ESP-IDF projects to supported microcontrollers.