Kinesis 0.1.0 – Native macOS Controls for the Meta Neural Band

What Kinesis 0.1.0 Does

Kinesis 0.1.0 provides native macOS control bindings for the Meta Neural Band, enabling swipe‑between‑desktops, Mission Control activation, music playback control, and pinch‑plus‑turn gestures for volume or brightness. The project is written in Swift, built with Xcode Swift 6, and distributes as a notarized DMG for macOS 14+ on both Apple Silicon and Intel.

Installation and First‑Run Steps

  1. Download the latest DMG from the GitHub releases page.
  2. Drag Kinesis.app into the Applications folder.
  3. Launch the app; macOS will request Bluetooth and Accessibility permissions – grant both.
  4. Open the Setup tab, locate the Neural Band, pair it, and enable the desired control mappings.
  5. Use the band: swipe left/right to change spaces, pinch‑turn to adjust volume/brightness, and press the band‑mounted button to open Mission Control.

Building from Source

# Clone the repository
git clone https://github.com/callbacked/kinesis.git
cd kinesis

# Build the app (requires Xcode with Swift 6)
./scripts/build.sh

# Open the built application
open dist/Kinesis.app

Note: Quit the running Kinesis app before rebuilding to avoid file‑lock errors.

Technical Architecture

  • Language: 94.6 % Swift, 4.1 % Shell scripts, 1.3 % JavaScript.
  • Core Logic: Directly reads EMG signals from the Meta Neural Band (originally prototyped in the neural‑band‑poc repo).
  • Gesture Mapping: Configurable mappings stored in the app’s preferences; users can customize which EMG patterns trigger which macOS actions.
  • Distribution: The DMG is notarized, satisfying macOS Gatekeeper requirements.

Community Feedback Highlights

"The neural band always looked more interesting than the glasses… I’d be interested in exploring if something like this could remove the need for a mouse or touching the screen." – geoah

"Kinesis is already a word and an Amazon product, might have better discoverability naming your project something unique and specific." – pluc

"Very interesting. So how open is the neural band to hack on? Do you expect this to be something to get patched out eventually if you update?" – f0e4c2f7

"The vibe coders have finally discovered their killer interface, and it is kind of cool, especially if integrated with voice inputs." – fidotron

These comments underline two recurring themes: the novelty of EMG‑based control and the desire for broader hardware compatibility and open‑source driver support.

Limitations and Open Questions

  • Hardware Availability: Meta sells the Neural Band only bundled with its AR glasses, limiting access for developers who do not own the full headset.
  • Future Compatibility: The repository does not currently document a roadmap for handling potential firmware updates from Meta that could break the EMG interface.
  • Naming Conflict: "Kinesis" is also a trademarked Amazon product and a well‑known mechanical‑keyboard brand, which may affect discoverability.

Outlook

Kinesis demonstrates that consumer‑grade EMG wearables can be repurposed for desktop interaction without proprietary SDKs. If Meta opens the band’s firmware or if third‑party EMG bracelets become affordable, the approach could evolve into a mainstream alternative input method, especially when combined with voice assistants or gaze‑tracking technologies.

Sources

Related

  • Dispatch
    Swift Diffusers for Mac 1.1 Release

    Hugging Face has released Diffusers for Mac version 1.1, a native open-source app that leverages Core ML to accelerate Stable Diffusion on Apple Silicon, offering up to 2x faster generation on certain hardware configurations.

  • Project
    digimata/quill

    A macOS‑only Swift tool that records mic and system audio locally, then transcribes both tracks on‑device using a Core ML speech model, producing speaker‑tagged transcripts without sending any data to the cloud.

  • Project
    OpenWhispr/openwhispr

    OpenWhispr is a cross‑platform desktop app for privacy‑first voice dictation, meeting transcription, and AI‑assistant interaction. It supports offline speech‑to‑text with Whisper‑based models or cloud services, offers speaker diarisation, note‑taking with semantic search, team collaboration, and a public API for developers.

  • Project
    remorses/usecomputer

    usecomputer is a cross‑platform CLI (and TypeScript library) that lets AI agents take screenshots, map coordinates, and perform mouse/keyboard actions (click, drag, type, scroll) on macOS, Linux and Windows. It uses a native Zig backend, provides a coord‑map system for accurate screen‑to‑image mapping, streams global input events via SSE, and supports Kitty Graphics Protocol for agent‑friendly image delivery.

  • Project
    AgentiLoop/Agent

    A native macOS app that provides an autonomous AI agent capable of controlling the desktop, managing codebases, and automating system tasks across 18 LLM providers.