HiveMind Remi

A small web GUI for testing connections to a hivemind-core server. It is built with the Remi framework, so the interface is a local web page driven entirely by Python. Fill in your server credentials, connect, and chat with your assistant from the browser.

Where it sits

HiveMind is a mesh. Satellite devices connect to a central hivemind-core server over an authenticated, encrypted protocol. Remi is a throwaway test satellite. It wraps the Python hivemind-bus-client in a GUI so you can check that a server and access key work, and watch utterances and spoken replies flow, without writing any code.

Remi web GUI (hivemind-bus-client)  ──encrypted──►  hivemind-core  ──►  OVOS / agent

Install

pip install hivemind-remi

Or from source:

git clone https://github.com/JarbasHiveMind/HiveMind-remi
cd HiveMind-remi
pip install .

Runtime dependencies (remi, hivemind-bus-client, ovos-bus-client, ovos-utils) are declared in pyproject.toml, the single source of truth for packaging. See docs/dependencies.md for the version policy.

Quickstart

1. Run a server and issue an access key

On the machine hosting the assistant, install and run hivemind-core:

hivemind-core add-client      # prints an access key + crypto/password key
hivemind-core listen --port 5678

2. Launch Remi

HiveMind-remi

Remi serves its GUI as a local web page and opens it in your browser.

3. Connect and chat

On the Connect tab, fill in:

Click Connect. When the status reads Connected to HiveMind!, switch to the Chat tab, type a message, and press Send. The assistant's spoken responses appear in the chat log.

Documentation

Full docs live in docs/:

• Installation
• Configuration: the Connect-tab fields explained
• Architecture: the remi web-GUI design and how it talks to hivemind-core
• Dependencies: version policy and the 2.x stack
• Running tests: the smoke suite and the HiveMind-side end-to-end suite

License

Apache 2.0. See LICENSE.