Chipwright

Build a production smart device from a single definition, visually, in VS Code.

Describe a device once, in plain terms, and Chipwright generates its firmware interface, cloud API, tests, and docs. This extension is the visual front end: design a device, watch it run as a live digital twin, generate real code, and get AI help, without leaving the editor.

Features

Visual device designer

Define what a device senses, controls, how it connects, and how it is powered, through a form instead of hand-written config. It reads and writes a plain device.yaml, validates as you type, and shows the artifacts your definition produces.

Live twin debugger

Run the device as a digital twin before you own any hardware. Stream live telemetry to a chart and inject sensor faults (stuck, failed, drifting) to see how the device behaves.

Generate and scaffold

One click writes every generated artifact (firmware interface, cloud API, tests, docs) to a generated/ folder next to your manifest. A second click scaffolds a starter firmware module wired to that interface and opens it, so "design" flows straight into "code." Generated output is regenerable; your scaffolded code is yours and is never overwritten.

AI assistant, bring your own key

A chat panel that proposes edits to your device and shows them as a diff. Every suggestion is checked against the real compiler before you see it, and nothing is applied until you click Apply. Use Anthropic, Gemini, any OpenAI-compatible endpoint, or a fully local model. Your key is stored in VS Code secret storage and sent only to the endpoint you configure.

Getting started

1. Install the extension and open the Chipwright view from the Activity Bar.
2. Create a device from a template, or open an existing device.yaml.
3. Design it in the Designer, then use Generate and Scaffold firmware from the output panel.
4. Switch to the Twin tab to run it and watch live telemetry.
5. Open the AI Assistant (sparkle icon) to edit the device from a prompt.

Requirements

• The designer, artifact generation, and the AI assistant work in any workspace.
• The twin debugger runs Chipwright's digital twin, which lives in the Chipwright source workspace; open that repository to use it. Building the twin needs a C toolchain (make plus gcc or clang).

Extension settings

The assistant's API key is set through the panel and stored in VS Code secret storage, never in settings.

Learn more

• Documentation and guided tour
• Report an issue
• Source code

Chipwright is created and maintained by Diego Regalado. Licensed under the Elastic License 2.0.