Thanks that's kind! My

play_special_effect

function has this description:

Show a lighting effect and play a sound effect, in order to add atmosphere and immersion to a story, scene, etc., or if you want to indicate a particular state that you're feeling, such as being sleepy, excited, celebrating, etc. Only use one special effect per sentence.

...and a single

effect_name

parameter with this description:

The name of the special effect to play. 
Valid values are:
- "magical_spell": When magical spells are cast.
- "lightning": Thunder and lightning.
- “rain": Raindrops falling.
- "rainbow": A rainbow of colors and mysterious sounds.

(I'm surprised there's no way to specify an "enum" of valid values for the string, so I'm just guessing that this is the best way to constrain it, but it does work okay). As far as switching the function from async to synchronous, I could give that a shot. I just don't want/need a pause in speech output, nor a "Request Start Message" to be spoken. Will try just putting a filler word in there. Thanks again!