OpenRayneo skill
OpenRayneo lets a local agent control RayNeo iO glasses over an HTTP API: notifications, teleprompter, todos, audio capture.
- 111 ★ on the repository
- Last commit
- Apache-2.0
Extended access
Why this level
- The bridge API can record audio from the glasses microphone and forward it.
- The API token gives control of the glasses and must stay local.
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add skills/openrayneoDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewExtended accessThe tool runs code or makes network calls and only reads external data.
- The bridge API can record audio from the glasses microphone and forward it.
- The API token gives control of the glasses and must stay local.
This is third-party code. Review the repository files before installing.
What it does
OpenRayneo is an unofficial companion app for RayNeo iO glasses on macOS and Android. The repository contains the `openrayneo-api` skill: connection guidance, request examples and a helper script, `scripts/request.py`. It lets a local agent work with an already running bridge but does not install the app or pair the glasses. The skill works through the HTTP API at `OPENRAYNEO_API_URL` with the `OPENRAYNEO_API_TOKEN` token: short notifications via `/v1/notifications`, a teleprompter with pause and scroll, captions, todos and weather, glasses audio recording, speech recognition and a virtual microphone. The script reaches only the explicitly configured local address, uses no proxy and does not retry requests on its own.
Who it is for. RayNeo iO owners and developers who want to connect Codex or another local agent to the glasses.
Good fit when
- An agent should send a short task-finished notification to the glasses.
- A long text should show as a teleprompter with pauses.
- Voice from the glasses should become text with the answer displayed.
- The glasses should act as a Codex microphone through a virtual device.
Not a fit when
- There are no RayNeo iO glasses or the bridge is not running.
- You need Bluetooth pairing: the HTTP API does not do it.
- Audio recording is not allowed and the agent might switch it on by mistake.
Example request
Check the connection to the glasses through OpenRayneo and send the notification "Build finished".Limitations
It is an unofficial project, needing no changes to the official app, but it is not supported by the manufacturer. Android navigation uses a fixed text grid, and arbitrary graphics and screen mirroring are not supported. The skill works with an already running bridge that must be set up per the app docs. The API token must not be printed or committed. The skill uses `uv run python`.
How to disable. Delete the `openrayneo-api` directory from the agent skills directory, close the bridge app and revoke the API token.
FAQ
How do I install the skill?
Copy `.agents/skills/openrayneo-api/` into the agent skills directory; for Codex that is `${CODEX_HOME:-$HOME/.codex}/skills/`.
What is needed to work?
A running bridge, the `OPENRAYNEO_API_URL` address and the `OPENRAYNEO_API_TOKEN` token in the environment.
Does the skill pair the glasses?
No, pairing is done by the app, and the skill uses an already running bridge.
What does `/health` show?
That the service is running. Connection readiness is confirmed by `bleReady: true` and `rfcommConnected: true`.
What else does the repository contain?
macOS and Android apps, a character video mode on a 26 by 7 grid and connection guides.
Discussion
No questions or reviews yet.
Related
SkillsAn MCP server that lets an agent see the screen and control macOS, Windows and Linux: screenshots, clicks, OCR and screen memory
A skill that wires Agent-S into OpenClaw: the agent hands over a task in words and Agent-S clicks through desktop windows itself.
A CLI that connects Claude Code or Codex to robots and simulators through ROS 2: the agent writes perception and control programs.
Turns one run of a task into a program an agent calls as a single tool, checking the record before and after the save
Open Computer Use
open-computer-use
An open MCP server for desktop control: the agent reads app state and clicks, types and scrolls on macOS, Linux and Windows
A local macOS CLI that lets an agent read the screen via OCR, click and type, with no cloud and no vision model
README in short
The README covers the macOS and Android apps, character video on a 26 by 7 grid and the distributable skill. It gives the copy path for the skill into an agent directory.