Progress on my OpenClaw HA Voice PE device! First pass used a proxy to pass the speech-to-text output into `openclaw agent` on the CLI, await a response, and then send it back to HA, which would use TTS to speak a response. Working on updating the proxy to use WebSocket for streaming responses.

Note February 14, 2026 at 5:12 PM

Jonathan's location at time of posting:

LaCour

stationary

-3.6 km/h

50%

Also Posted To

Mastodon BlueSky