Though I thought I confirmed the consonants of my conlang (ѲКМНПҀСТФЦЧШ), eventually, I found that I'm in disfavor of the syllable composition of CVC if the "C"s were meant to be arbitrary consonants. So while keeping CVC, I decided to add more consonants to act as codas. Namely approximants, or ВЙЛР to be exact. To give their exact pronunciation in IPA, Ѳ [θ], К [k], М [m], Н [n], П [p], Ҁ [ʡ], С [s], Т [t], Ф [f], Ц [t͜s], Ч [ʈ͡ʂ~t͜ɕ], Ш [ʂ~ɕ], В [w~v], Й [j~ʝ], Л [l], and Р [ɻ].
I wanted to take this further. Namely, ѲКМНПҀСТФЦЧШ should act only as onsets, and ВЙЛР should act only as codas. But is it plausible that onsets and codas are entirely disjoint sets?
Take Korean for comparison. It has ㄱㄲㄴㄷㄸㄹㅁㅂㅃㅅㅆㅈㅉㅊㅋㅌㅍㅎ as onsets, and ㄱㄴㄷㄹㅁㅂㅇ as codas; there is an intersection.