Build MicroPython hardware projects from a single sentence. Describe what you
want to build — Blockless picks a board, selects and wires the parts, generates
deployable MicroPython code, and walks it onto your device, all inside a VS Code
side panel.
You type:"A temperature display that warns when it's too hot."
Blockless gives you: a matched board, a parts list with a wiring diagram
you can read, and audited MicroPython code ready to flash.
What it does
Plain language to a real project. A sentence becomes a board choice, a
wired parts list, and a working MicroPython project.
Real component intelligence. Parts, drivers, and pinouts come from a
curated package catalog — not APIs guessed by a model.
Wiring you can read. Every build shows the parts and a wiring diagram
derived from the project manifest.
Generate, audit, deploy. Code is audited for unsafe imports, then flashed
and run on a connected board — with a confirmation checkpoint before anything
touches hardware.
Speaks your language. The UI follows the language you type in.
How it works
Describe what you want to build, in plain language.
Review the plan — board, parts, and wiring, laid out before any code.
Generate deployable MicroPython, audited for unsafe imports.
Deploy to a connected device over USB, with a confirmation checkpoint
first.
Getting started
Install the extension and open the Blockless view from the activity bar.
Sign in with GitHub when prompted (used for your free daily credits).
Describe what you want to build and follow the build plan.
To deploy to a real board, connect a MicroPython device over USB — the
extension runs the local device helper for you.
Requirements
A GitHub account for sign-in and credits.
For deployment: a MicroPython-capable board (e.g. ESP32) connected over USB.
Settings
mpyhw.apiBaseUrl — override the backend API URL (e.g. a self-hosted
backend). Leave blank to use the default hosted backend.
mpyhw.pythonPath — developer override for the fallback local runner.
mpyhw.pipIndexUrl — developer override for fallback runner dependency setup.
Privacy
Your prompts and generated code are sent to the Blockless backend, which calls a
large language model to plan and generate your project. Telemetry is sanitized —
prompts and code are hashed, not stored verbatim.