build Macro for android

I have a voice assistant app (UTTER) on my 7.1.1 It has voice recognition and responds verbally to my request, similarly to OK google. When it responds, I want to mute the bluetooth phone output to mask the recognition beep sound lasts about 300ms, played before each response. Can I detect the begining of an audio stream and mute the bluetooth output for the time the sound is playing (like 300ms), then unmute the bluetooth. Alternatively, the audio stream can be muted for the first 300ms, if that is more efficient. UTTER has a shake to wake feature, so each time i want to talk to UTTER, I shake the phone and UTTER responds with "may I help you", with NO beep sound, so i don't want to mute the beginning of that verbal response. Another feature is to detect the phone shake and disable the macro mute effect for UTTER's "may i help you" response and then enable mute feature for next response (like giving the time), or instead of detecting shake, detect first verbal response (or set a counter to 0) and disable mute for all odd number verbal responses.