Reconstructions

PCS
*ˁagal-at- - wagon (Kogan 2015: 207)
PCS
*ˁḳr - to be barren (Kogan 2015: 182)
probably related to *ˁḳr ‘to uproot,’attested in Hebrew and Aramaic
PCS
*ˁlm - (to) sign (Kogan 2015:40, note 94)
Contra HALOT 834, there is hardly any connection with Hbr. ˁlm ‘to be hidden, concealed.’
PCS
*ˁimm(-a) - together with (Arakelova 2001:41; Huehnergard 2005:190; Hasselbach–Huehnergard 2007:420; Kogan 2015: 182)

Lipiński (1997:465) derives this preposition from the PS verbal root ‘to be common’. More cautious treatments can be found in Gesenius 1910:297 (”ˁim- (connexion ?) with”) and Tropper 2000:763 (”ˁm /ˁimma/ (Subst. /ˁimm-/? “Gemeinschaft” im Ak.)”).

PCS
*ˁamm- - relatives, clan, people (Kogan 2011: 235; Kogan 2015: 182)
probably a semantic extension of the kinship term *ˁamm- ‘grandfather, ancestor’
PCS
*ˁamal- - work, toil, labor (Kogan 2008b:101)
PCS
*ˁinab- - grape (Kogan 2015: 187)
PCS
*ˁny, ˁnw - to be distressed, humbled (Kogan 2015: 183)
cf perhaps Tna ˁanäwä to collapse, to crumble, to fall
PCS
*ˁar(ā)d- - wild ass (SED II No. 37, Kogan 2011: 208)
PCS
*ˁVŝb- - grass (Fronzaroli 1968: 274, 289, 299; Kogan 2011: 200; Kogan 2015: 183)
The only possible trace of this root outside CS can be found in Akk. ešēbu ‘to grow luxuriantly’. Its relationship to the Akkadian plant name išbabtu is considerably less certain because of the peculiar morphological shape of the Akkadian word.
Sab. ˁs₃bt ‘pastureland’ and Qat. ˁs₃b ‘crops, produce’, often compared to this root (Fronzaroli 1968:289), can hardly be related in view of the phonological irregularity (ESA s₃ does not normally correspond to PS ŝ) (Kagan 2015: 183)