Inference of cell-cell communication (CCC) from single-cell RNA-sequencing data is a powerful technique to uncover putative axes of multicellular coordination, yet existing methods perform this analysis at the level of the cell type or cluster, discarding single-cell level information. Here we present Scriabin â€" a flexible and scalable framework for comparative analysis of CCC at single-cell resolution. We leverage multiple published datasets to show that Scriabin recovers expected CCC edges and use spatial transcriptomic data, genetic perturbation screens, and direct experimental manipulation of receptor-ligand interactions to validate that the recovered edges are biologically meaningful. We then apply Scriabin to uncover co-expressed programs of CCC from atlas-scale datasets, validating known communication pathways required for maintaining the intestinal stem cell niche and revealing species-specific communication pathways. Finally, we utilize single-cell communication networks calculated using Scriabin to follow communication pathways that operate between timepoints in longitudinal datasets, highlighting bystander cells as important initiators of inflammatory reactions in acute SARS-CoV-2 infection. Our approach represents a broadly applicable strategy to leverage single-cell resolution data maximally toward uncovering CCC circuitry and rich niche-phenotype relationships in health and disease.