This document used to describe an Amazon Alexa setup procedure (Echo device → Cloudflare tunnel → YWC). That approach has been retired in favour of a much simpler local solution.
Voice control in YWC now uses Windows' built-in speech recognition (SAPI 5 /
System.Speech) running locally on the same PC as YWC. A microphone connected to the PC is the only hardware requirement — no Echo, no Alexa Skill, no domain, no Cloudflare account, no public endpoint. Recognised audio never leaves your computer.Voice control v1 shipped in v2.4.0-pre1 (2026-06-24) and is part of the current pre-release v2.4.0-pre2.
- USER_MANUAL.md §17 Voice Control — full description of what voice does, the supported command set, how to enable it, the en-GB Windows speech-pack install step, and troubleshooting.
- USER_MANUAL.md §5.16 Voice Announcements is the separate feature that makes YWC speak to you (band/mode/TX cues). Voice Control (§17) is you speaking to YWC. The §17 introduction includes a callout disambiguating the two.
The original 220-line setup procedure (covering Amazon Developer Console, Cloudflare DNS migration, Alexa Skill JSON, cloudflared install, signature verification, etc.) is preserved on the feature/alexa-voice-control branch:
git checkout feature/alexa-voice-control -- VOICE_CONTROL.md
The proof of concept on that branch did work end-to-end including signature verification — a real Echo device successfully drove a local YWC over a Cloudflare tunnel — but the setup commitment (well over an hour of fiddly configuration per user) was too high for the audience. The local SAPI 5 approach takes about two clicks in YWC Settings, runs offline, and is the supported voice-control path going forward.
The branch is retained for reference in case any future work needs to revisit the Alexa route or borrow patterns from the signature-verification code there.