Okay, so to use your voicebank you need to run it through either presamp or autocvvc, for which you may need to create a presamp.ini first. CVVC in general is better off used in OpenUtau or using the
PhonemizeronUTAU plug in just for sanity's sake but it's up to you.
But also it not being complete will not help of course haha. IMHO using the base oto is a better idea bc it means you dont have to manually alias which is very taxing for a voicebank type that's already hard to test unless you're already quite familiar with UTAU or are using OpenUtau.
Using the base oto, OpenUtau, and vlabeler I got a "complete" and usable alias situation so I could test:

Only oto'd these lines in full as a test to see if everything was fine, it is, no overlap blips... So you just gotta keep going, and make sure you're using something that lets you actually use a CVVC bank.
A basic visual in case you wish to continue manually aliasing, before and after of your VCs:


As you can see, the preutterance needs to be at the start of the consonant rather than the middle or end to prevent early starts or strange overlapping. The overlap should be just before the previous vowel begins to taper off in the waveform, or spectrogram (my screenshots are in vlabeler, but you can get a spectrogram view in any otoing program).
Additionally, VCs will ALWAYS be named "n f" and not "ん f" as all solutions for VCCV use a standardized "romaji prior vowel, next consonant" format to match VCVs "romaji prior vowel, hiragana CV" format. You do, though, need hiragana CVs. Your own aliases also miss out on the beginning (- CV) sounds, which are part of why you say the syllable twice-- That way you have - ka for the start of phrases, as well as ka for after a VC. They have different needs, so it's important and possibly part of why you're encountering issues.
Another example, this time how i would recommend entries for your go-go file be done.


Some reading for ya that should help with some basics that seem to be throwing off your V C otos in particular, though your CVs also need some brushup (preutterance always directly at the end of the consonant and start of the vowel btw):
https://yinsototutorial.weebly.com/9-cvvc-japanese.html
https://utaforum.net/resources/anatomy-of-the-oto.321/
https://github.com/stakira/OpenUtau/wiki/Voicebank-development
I'm blunt, but I hope this doesn't come across as harsh. It's hard to tell what you need and how best to convey it when you're giving this little information so i just have to generalise.