TL;DR: This appears to be the most striking example known of dental convergence and possible competition between a mammal and a lizard, which in the long run worked out better for the lizards.
Abstract: Extinct species of Malleodectes gen. nov. from Middle to Late Miocene deposits of the Riversleigh World Heritage Area, northwestern Queensland, Australia are enigmatic, highly specialized, probably snail-eating marsupials. Dentally, they closely resemble a bizarre group of living heterodont, wet forest scincid lizards from Australia (Cyclodomorphus) that may well have outcompeted them as snail-eaters when the closed forests of central Australia began to decline. Although there are scincids known from the same Miocene deposits at Riversleigh, these are relatively plesiomorphic, generalized feeders. This appears to be the most striking example known of dental convergence and possible competition between a mammal and a lizard, which in the long run worked out better for the lizards.
TL;DR: Phylogenetic analysis of 62 craniodental characters places Malleodectes within Dasyuromorphia albeit with weak support and without indication of specific relationships to any of the three established families (Dasyuridae, Myrmecobiidae and Thylacinidae).
Abstract: A new specimen of the bizarrely specialised Malleodectes mirabilis from middle Miocene deposits in the Riversleigh World Heritage Area provides the first and only information about the molar dentition of this strange group of extinct marsupials. Apart from striking autapomorphies such as the enormous P3, other dental features such as stylar cusp D being larger than B suggest it belongs in the Order Dasyuromorphia. Phylogenetic analysis of 62 craniodental characters places Malleodectes within Dasyuromorphia albeit with weak support and without indication of specific relationships to any of the three established families (Dasyuridae, Myrmecobiidae and Thylacinidae). Accordingly we have allocated Malleodectes to the new family, Malleodectidae. Some features suggest potential links to previously named dasyuromorphians from Riversleigh (e.g., Ganbulanyi) but these are too poorly known to test this possibility. Although the original interpretation of a steeply declining molar row in Malleodectes can be rejected, it continues to seem likely that malleodectids specialised on snails but probably also consumed a wider range of prey items including small vertebrates. Whatever their actual diet, malleodectids appear to have filled a niche in Australia’s rainforests that has not been occupied by any other mammal group anywhere in the world from the Miocene onwards.