Biological feedback mechanisms exert precise control over the initiation and termination of molecular self-assembly in response to environmental stimuli, while minimizing the formation and propagation of defects through self-repair processes. Peptide amphiphile (PA) molecules can self-assemble at physiological conditions to form supramolecular nanostructures that structurally and functionally resemble the nanofibrous proteins of the extracellular matrix (ECM), and their ability to reconfigure themselves in response to external stimuli is crucial for the design of intelligent systems. In this thesis, we investigated the real-time self-assembly, deformation, and self-healing of ECM-mimetic PA nanofibers in aqueous solution by using a force-stabilizing double-pass scanning AFM imaging method to disrupt the self-assembled peptide nanofibers in a force-dependent manner. We showed that nanofiber damage occurs at tip forces exceeding 1 nN, and that the damaged fibers subsequently recover under sub-nN tip forces. Fiber ends occasionally failed to reconnect following breakage and continue to grow as two individual nanofibers. Energy minimization calculations of nanofibers with increasing cross-sectional ellipticity (corresponding to varying levels of tip-induced fiber deformation) supported our observations, with high-ellipticity nanofibers exhibiting lower stability compared to their non-deformed counterparts. As a result, tip-mediated mechanical forces can provide an effective means of altering nanofiber integrity and visualizing the self-recovery of PA assemblies.