Animal chromosomes are partitioned into contact domains. Pathogenic domain disruptions can result from chromosomal rearrangements or perturbation of architectural factors. However, such broad-scale alterations are insufficient to define the minimal requirements for domain formation. Moreover, to what extent domains can be engineered is only beginning to be explored. In an attempt to create contact domains, we inserted a 2-kb DNA sequence underlying a tissue-invariant domain boundary—containing a CTCF binding site (CBS) and a transcription start site (TSS)—into 16 ectopic loci across 11 chromosomes, and characterized its architectural impact. Depending on local constraints, this fragment variably formed new domains, partitioned existing ones, altered compartmentalization, and initiated contacts reflective of chromatin loop extrusion. Deletions of the CBS or the TSS individually or in combination within inserts revealed their distinct contributions to genome folding. Altogether, short DNA insertions can suffice to shape the spatial genome in a manner influenced by chromatin context. Additionally, we frame our findings in the context of recent advances in experimentally creating chromatin loops, contact domains, boundaries, and compartments. Furthermore, we explore parallels between this emerging theme and natural evolution of mammalian genomes with increasing architectural complexity. Finally, we provide a perspective on how insights arising from recent gain-of-function studies may inform future endeavors towards engineering the three-dimensional (3D) genome.