hivemind-webspeech

hivemind-webspeech is a browser speech client for HiveMind. It captures your microphone, runs voice activity detection (VAD) in the page, and streams each spoken utterance as base64 audio to a HiveMind hub over an encrypted WebSocket. The hub handles speech-to-text, intent, skills, and the reply.

No install, no Python, no audio drivers — a web page and a microphone.

The satellite spectrum

HiveMind satellites differ by where the work happens. This client sits at the thin end (browser variant): microphone capture and VAD run locally, everything else runs on the hub.

Client	Runs locally	Runs on the hub
HiveMind-cli	nothing (text only)	STT · TTS · intent · skills
hivemind-webspeech ← you are here	microphone · VAD	STT · TTS · intent · skills
hivemind-mic-satellite	microphone · VAD	STT · TTS · intent · skills
HiveMind-voice-relay	mic · VAD · wake-word	STT · TTS · intent · skills
HiveMind-voice-sat	mic · VAD · wake-word · STT · TTS	intent · skills

hivemind-webspeech is the browser counterpart of the native hivemind-mic-satellite: the same division of labour, a web page instead of a Python process, built on HiveMind-js rather than hivemind-websocket-client.

Documentation pages

Page	What it covers
Getting started	Pair, allow the audio message, open the page, speak
Configuration	Credential fields, ports, the TLS / mixed-content rule, hub requirements
Audio pipeline	How mic → VAD → WAV → base64 → bus → reply works on the wire
Troubleshooting	Connection, TLS, audio, and VAD problems

Quick reference

# on the hub:
hivemind-core add-client                              # → Access Key + Password
hivemind-core allow-msg "recognizer_loop:b64_audio"   # permit audio over the bus

Then open the demo (or your own build), enter IP / port / access key / password, click CONNECT, then Start VAD, and speak.