It is commonly presumed that abiotic membranes were colonized by proteins later on. Yet, hydrophobic peptides could have formed primordial protein-dominated membranes on their own. In a metabolism-first context, "autocatalytically closed" sets of statistical peptides could organize a self-maintaining protometabolism, assisted by an unfolding set of ribotide-related cofactors. Pairwise complementary ribotide cofactors may have formed docking guides for stochastic peptide formation, before replicating RNA emerged from this subset. Tidally recurring wet-drying cycles and an early onset of photosynthetic activities are considered most likely to meet the thermodynamic requirements. Conceivably, the earliest peptide-dominated vesicles were engaged in light harvesting, together with isoprenoid-tethered pigments, rather than providing an external boundary. Early on, the bulk of prebiotic organic matter can have formed a contiguous layer covering the mineral sediment, held in place by colloidal coherence of a hydrogel matrix. This unconventional scenario assumes a late onset of cellular individualization - perhaps from within, resembling endosporogenesis.