We present a novel method to differentiate stream-like and shell-like tidal remnants of stellar systems in galactic halos using the density-based approach of the clustering algorithm AstroLink. While previous studies lean on observation, phase-space, and action-space based criteria for stream and shell determination, we introduce AstroLink’s ordered-density plot and cluster identification as a viable tool for classification. For a given data set, the AstroLink ordered-density plot reveals the density-based hierarchical clustering structure from which the resultant clusters are identified as being statistically significant overdensities. Using simulations of sub-halo disruptions in an external potential to generate samples of tidal structures, we find that the curvature of the ordered-density plot is positive for stream-like structures and negative for shell-like structures. Comparisons with more standard classification techniques reveal strong agreement on which structures typically fit into stream-like and shell-like categories. Furthermore, we investigate the properties of clustered stream and shell samples in radial phase space and energy-angle space. Given the sensitivity of stellar tidal structures to their host dark matter halos, the identification and subsequent classification of these structures provide exciting avenues of investigation in galactic evolution dynamics and dark matter structure formation.