This might be helpful.

https://psychologicalstatistics.blogspot.com/2021/08/i-will-not-ever-never-run-manova.html

Generally MANOVA is followed up with univariate ANOVA anyway and there's no good reason to prefer MANOVA. It's atheoretical (I.e., you analyse not three DVs but some atheoretical linear combination of them) and it doesn't necessarily protect FW error. I'd just run the three ANOVAs and if I was concerned about FW error control use a correction like the Hochberg correction.