Wound healing is a complex orchestration of chemical signalling, cellular activity, and local changes in extracellular matrix (ECM). These processes lead to emergent macroscopic changes: replacement of the fibrin clot by collagen-dense and stiff scar tissue and contraction of the wound. Despite knowing the elements of the wound healing cascade, there is still a lack of predictive understanding and therefore ability to control the process to avoid pathological scarring. Computational models are ideal to dissect fundamental relationships and generate testable hypotheses that can improve wound treatment paradigms in the near future. We propose a theoretical and computational model of wound healing with emphasis on the mechanobiological and biomechanical couplings. The model accounts for two inflammatory cytokines that obey reaction diffusion. (Myo)Fibroblast population density is also modelled. (Myo)Fibroblasts directly couple to tissue mechanics by depositing collagen and exerting contractile stress, and their activity is regulated by mechanobiological feedback (local state of stress/strain). The model captures increased collagen density which surpasses physiological values and is unable to return to baseline by day 21. The wounded tissue, which is initially soft and has a relatively linear stress-strain response, becomes nonlinear and stiffer than unwounded tissue. With the calibrated model based on murine data, we test two hypotheses. First, we test whether collagen content alone can explain the observed tissue mechanics and we conclude that collagen content alone is unable to capture the stiffening of the scar tissue. Instead, we pose that secondary mechanisms such as collagen crosslinking are required to match observations. Second, we test if strain-driven or stiffness-driven coupling can explain hypertrophic scarring. We find that stiffness-driven coupling is required, and this coupling produces a bi-stable system in which fibrosis is a possible steady state.