Salut1231/wyoming-voice-match ? reverse-engineered prompt

Build me a small Python service for Home Assistant that sits between my microphone and speech to text, checks whether it sounds like the right person, cleans up the audio, and then forwards only the good voice command onward. I want it to help reduce false triggers from TV, radio, and other people talking nearby, while still working smoothly with Wyoming protocol style voice setups. Please make it easy to run in Docker, with a normal and CPU friendly setup, and include a simple way to enroll a voice so the system can recognize an approved speaker. It should be straightforward to configure, log what it is doing, and be usable from a local network address and port. If you need to, look up current docs online for anything unclear, but keep the setup simple and practical.