This is awesome!!!

Letters with an acute or double acute accent are pronounced as a rule about twice as long as letters with a caron or no diacritic. This is represented in the phonetic transcription using either the ː symbol or by simply writing the letter twice. Note, however, that there is some amount of dialectal variation in how the letters are pronounced; this is the "anchorman's speech". Also, the voiceless stops (K, P, T) don't necessarily have as much aspiration as they often have in English.

A Á /a, aː/ — somewhere between the vowel in "hat" and "bra" in a typical, like, Ohio dialect

B B́ /b, bb/ — a regular B sound like in "butler"

C Ć /ts, ccɕ/ — C is like the end of "hats"; Ć is like that but you use the middle of the tongue instead of the tip (palatalization)

Č C̋ /ʈʂ, ʈʈʂ/ — close to the ⟨ch⟩ in "change", except the tongue is curled backwards (retroflex)

D D́ /d, ɟɟ/ — D is a regular D sound like in "delicate"; D́ is a palatalized version

E É /e, eː/ — the vowel in "bed" in an Australian or New Zealand dialect

F F́ /f, fː/ — a regular F sound like in "firefighter"

G Ǵ /g, gg/ — a regular hard G sound as in "goose"

H H́ /h~x, xː/ — the short version of this letter is pronounced with a regular H sound like "hello" at the start of words, and is otherwise pronounced like the H of a stereotypical Russian accent, or like the ⟨ch⟩ in the Scottish English pronunciation of "loch"

I Í /i, iː/ — the vowel in the word "bit" in an Australian dialect, or like the vowel in "beet" in most English dialects

J J́ /ɖʐ, ɖɖʐ/ — the J in "juice" but retroflex

K Ḱ /k, kk/ — a regular K sound like in "kid"

L Ĺ /l, ʎː/ — L is a regular L sound like in "lemon"; Ĺ is a palatalized L, a bit like the ⟨lli⟩ in "million"

M Ḿ /m, mː/ — a regular M sound like in "mother"

N Ń /n, ɲː/ — N is a regular N sound like in "novel"; Ń is a palatalized N, a bit like the ⟨ny⟩ in "canyon"

Ň N̋ /ŋ, ŋː/ — like the ⟨ng⟩ in "singing" in most dialects

O Ó /o, oː/ — like the vowel in "yawn" in RP or "bro" in Indian English

Ǒ Ő /ø, øː/ — as said, like the vowel in "bird" in New Zealand and South African English

P Ṕ /p, pp/ — a regular P sound like in "pecan pie"

R Ŕ /ɾ, rʲː/ — R is like the tapped T or D in General American "latter ladder"; Ŕ is like the trilled R of Scottish and Welsh English, or the ⟨-d it⟩ in some AAVE realizations of "speed it up", except that the trilled Ŕ also involves, I want to say, raising the body of the tongue while trilling until it becomes impossible to trill any longer. It's a tricky sound to make but it's found in e.g. Russian царь (tsar').

S Ś /s, ɕː/ — S is a regular S sound like in "sight", Ś is a palatalized S sound which sounds a bit like a sharper "shush" sound.

Š S̋ /ʂ, ʂː/ — like the ⟨sh⟩ in "ship" except retroflex.

T T́ /t, cc/ — a regular T sound like "tomorrow" and its palatal equivalent.

U Ú /u, uː/ — like the vowel in Australian English "book", also similar to the vowel in "boot" in most accents but further back in the mouth ※may be realized as a W sound when next to another vowel

V V́ /v, vː/ — like a regular V sound like in "vendetta", except when used as the construct state suffix, particular before a consonant other than Y, W, or V, in which case the -v suffix is commonly read with a W sound

W Ẃ /w, wː/ — a regular W sound like in "water"

X X́ /dz, ɟɟʑ/ — X is like the ⟨ds⟩ in "pads"; X́ is the palatalized equivalent

Y Ý /j, jː/ — like a regular yod sound like in the word "yarn", except in a cluster of /jwC/ (C = any consonant), in which case the /jw/ will fuse into /ɥ/, which is like a yod sound with rounded lips.

Z Ź /z, ʑː/ — a regular Z as in "zebra" and its palatalized equivalent

Ž Z̋ /ʐ, ʐː/ — like the J in "j'accuse" except retroflex

No one else has got the other two yet, you wanna have another go?

Alright so my conlang's a bit of a mess so you'll have to forgive me for some of these things, but I'll do my best to setsumei suru.

CONS is construct state, which means that it's close to の but sort of "in reverse". This is to say, when a word is put in the construct state, it means that the word is modified by the following words, rather than modifying the following words itself. This is a feature most associated with Semitic languages — the go-to example is {malika|queen} vs {malikat|queen-CONS} {sabaʾ|Sheba} for "Queen of Sheba" in Arabic — but the construct state is also found in a number of other languages, Western Micronesian languages have them too for instance. So "kyertev _____" really means "a thing loved by _____", and when this phrase is the subject of a verb, it really means "_____ likes to..."

—Oh, I should clarify, the literal meaning of kyerte is just "a heart", a word with which it Definitely Isn't Cognate. So basically if we're translating things even more literally it's maybe more like "Rin's heart goes alone..." for "Rin likes to go alone..."

As for the other features:

1. Not necessarily obligatory, or at least I don't think of it as obligatory, but you're on the right track to think that kav anskagunska he is an idiomatic placeholder — it's really just the equivalent to "one day". Sometimes I question whether not marking tense in verbal inflections actually was such a good idea after all!
2. The lemmata for those words are an and gun, and yéne so yegune was really just supposed to mean "the two of them together", I wasn't necessarily thinking consciously about which character was yéne and which was yegune. There's no form of obviation in this lang in any case, and there's a healthy class of pronouns including interrogative pronouns, indefinite pronouns, personal pronouns, demonstratives, among others, all the classics.
3. The pronoun šo is used as a complementizer and relative pronoun (very European!), and can in general be used as a pronoun pointing to the previous clause, which is the meaning of šo in the phrase so šo. This phrase so šo most often means "thereafter" but may also just mean "as well", because my lang is perhaps a bit more vibes-based than it should be.

Then as for the orthography, Ňň represents the /ŋ/ sound (modeled on Turkmen), Šš, Žž, and Čč represent /ʂ/ /ʐ/ and /ʈʂ/ respectively (these three are indeed modeled on Slavic languages like Serbo-Croatian), and then Ǒǒ represents a vowel in the vicinity of /ø/ for really no other reason than that it would feel "inelegant" to suddenly use a different diacritic for that letter. The doubled versions of these caroned letters replace the caron with a double acute accent, so N̋ /ŋː/ S̋ /ʂː/ Z̋ /ʐː/ C̋ /ʈʈʂ/ Ő /øː/. Non-caroned letters all double by just adding a single acute above them.

The letters Aa, Bb, Ee, Ff, Gg, Ii, Kk, Mm, Oo, Pp, Ww are really pretty boring. They make exactly the sounds you'd expect from the IPA. Uu is normally read as /u/ but if next to another vowel may become /w/. Vv normally makes a /v/ sound, but can make a /w/ sound when used as the construct state suffix — officially only before consonants other than glides or another V, but in practice I often end up just reading the construct state suffix as /w/ almost always regardless of what comes after it. Yy normally makes a /j/ sound, but when followed by the aforementioned preconsonantal construct state suffix, /jw/ will fuse into /ɥ/.

The letter Qq is not normally used. The letter Hh makes a /h/ at the start of a word but a /x/ elsewhere. The letter Jj makes a /ɖʐ/ sound.

The remaining letters are the alveolars: Dd, Ll, Nn, Rr, Ss, Tt, Zz make the sounds you'd expect, Cc and Xx make /ts/ and /dz/ sounds respectively, and all of these when doubled will palatalize: D́d́ /ɟɟ/ Ĺĺ /ʎː/ Ńń /ɲː/ Ŕŕ /rʲː/ Śś /ɕː/ T́t́ /cc/ Źź /ʑː/ Ćć /ccɕ/ X́x́ /ɟɟʑ/