# openwyzeband

> Agent-facing counterpart to the [human project page](/projects/openwyzeband/).

## Record metadata

- Record: 035
- Slug: openwyzeband
- Domain: Hardware
- Domain code: HW
- Type: Reverse engineering
- Status: Paused
- Period: 2026
- Portfolio role: Hardware experiment
- Publication state: Public device-owned research
- Case-study readiness: Pairing blocked
- Compendium edition: 0.4

## Summary

An evidence-driven BLE toolkit for understanding and eventually controlling a personally owned Wyze Band.

## Overview

The Wyze Band contains a display and sensors that remain useful even when the vendor's software no longer provides the openness an owner wants. openwyzeband explores whether that hardware can be understood through careful Bluetooth evidence rather than guesswork.

A Bluetooth Low Energy research toolkit that scans the band, enumerates GATT services, records reads, writes, and notifications, and supports controlled protocol experiments.

Purpose: Create an open, documented path to the device's display and live sensors without guessing at unsafe commands.

## The problem behind the project

The closed device has useful sensors and a display but no durable open interface. Careful observation could make the hardware reusable after vendor software stops serving it.

Reverse engineering invites premature conclusions: a writable characteristic is not automatically safe, and one observed packet is not automatically a command. The project therefore treats logging and controlled experiments as the primary instruments.

Device owners, hardware hackers, and open-device researchers may benefit. Unsafe commands could damage a device or violate consent if used on hardware someone does not own.

## How it took shape

BLE discovery, GATT enumeration, structured logging, notification capture, and controlled experiment notes designed to turn observations into typed capabilities.

The toolkit scans for the owned device, enumerates GATT services, records reads, writes, and notifications, and supports experiments that change one variable at a time. The eventual goal is typed control of the display and direct live access to sensor data.

Josiah set the open-control goal, supplied the owned device, and directed an evidence-first agent workflow rather than blind protocol guessing.

The toolkit and protocol-research structure exist, but the band is currently having trouble pairing with Josiah's phone.

## What the project means now

That goal remains incomplete because the band is currently failing to pair reliably with Josiah's phone. Preserving the blocker is important: open control is a research direction, not a capability the current artifact has already delivered.

The project is paused at a pairing problem and does not yet offer full display control or reliable live sensor access.

Reverse engineering advances faster when every interaction is logged as evidence and experiments change one thing at a time.

Resolve pairing, recover a known-good baseline, then map display and sensor behavior through controlled device-owned tests.

## Publication and interpretation notes

- Current classification: Paused
- Portfolio readiness: Pairing blocked
- Publication boundary: Public device-owned research

## Additional agent context

The eventual goal is open display control plus live sensor data; those capabilities are not yet complete.

## Related project records

- [iPhone Camera Bridge](/projects/iphone-camera-bridge/llm/) — A local WebRTC bridge that lets a Mac-based agent use an iPhone as a user-controlled camera.
- [Agent Interface Experiments](/projects/agent-interface-experiments/llm/) — A family of experiments that let AI agents see, touch, navigate, and manipulate interfaces beyond a text-only shell.

## Navigation

- [Complete project index](/projects/llm/)
- [Human version of this record](/projects/openwyzeband/)
- [About Josiah's working method](/about/llm/)
- [Agent discovery map](/llms.txt)