móyanas bo
3F-sleep-INF be-3F
mu-Qyan-ahs *p-[m]uQ

móyanas apó
[ 3F-sleep-INF ] NEG-be-3F
mu-Qyan-ahs a-p-[m]uQ

amóyanas bo
[ NEG-3F-sleep-INF ] be-3F

or the [w] might be merely epenthetic, i.e. an original
*a-u-har- (with Null prefix u-)
underwent regular loss of intervocalic /h/ to yield
*aú-ar- > oóar-

Original root *haQr-
gave tonic hér- and atonic -[w]ar-

analogous to 'say' *saQn-
which gave sén- and -san-

Je le déclare,
vous êtes des dieux,
vous êtes tous des fils du Très-Haut

("tous" here as [tus], I reckon, not *tu)

elsewhere ψ82

CORRECTION
There’s one root with initial cr:
a-crú ‘come (neg.)’

whose aorist is
ddumóomo, ddíimi etc.

*Th-
*T-
*D-

9. missed this one! here's more contrived tongue twisters idk what else i have:

bʕūd bʕīd bʕūd bʕīd bʕūd bʕīd bʕūd bʕīd bʕūd
"i will return on a holiday with a faraway oud; i again repeat: with a faraway oud i will return"

bʕīd bʕīd bʕīd
"i repeat & repeat & repeat"/"i repeat on a faraway holiday"

1 "-pi," 2 "bi-"
compare modern /bi, api/
pres. copula, nonhuman subj.

Some assumptions I'm working with:
- mid vowels are derived from high vowels (as noted, here both are adjacent to <h>)
- initial stops are voiced; voiceless ones barred from that pos'n
- zero initials reflect word boundaries
- at this stage vowel sequences were not tolerated (so <-i> = /ỵ/)

CORRECTION
line 2 replace "ng" with [ŋ]

I've moved the /l/ of to the following word; -lt- is a common sequence in native #Burushaski lexemes

/napan/ or /napang/ looks like a converb (participle);

/ku-/ of /kuỵ/ recalls the 'thou' 2s prefix gu- (basic form ku-)

3rd attempt.
Mid vowels only adjacent to h (uvular). Common final ŋ (> -mu(-c)) maybe a noun ending; clitics begin w/ r

χo napa[ŋ]-ri ltilpi
bu biti latitasing
un up χaŋ-paŋ-ril
uppisu baŋ-ri
žeχalpaỵmaćaŋ
kuỵ daŋ-ratti