Compelling geoscience evidence has heightened appreciation of abnormal (supra-hydrostatic, sub-/quasi-/supra-lithostatic) pore pressure and multi-mode (brittle-shear) fault rupture leading to seismicity. However, preparation and rupture nucleation are yet to be adequately constrained and quantitatively modelled. This challenges crucial schemes, notably physics-based earthquake forecasting/prediction. In this multidisciplinary study, transitions and associated critical points, linking pre-exisiting fluid overpressure in fault patches, temporal crustal stress and sequential brittle-shear fault failure, were considered. In preparation, tectonic loading was accompanied by processes such as off-fault yielding, permeability enhancement and foreshocks that facilitate local/regional stress relaxation. Subsequent equalization of stress and pre-exisiting local overpressure triggers hydraulic fracturing that destabilizes major asperities. Almost instant shear failure follows with spatially varying rupture velocity/intensity of frictional slip because of localized asperity stress and syn-slip fluid-/melt-driven fracturing/dilation and lubrication. Based on aforementioned critical points, quantitative modelling associated stress drop with evolution of stress and pore pressure. Equations for time to onset of stress relaxation and time to rupture were derived using fluid flow and viscoelastic models. Seismic moment was estimated with classical seismological relations after modifications accounting for less surface area during frictional slip. For retrospective testing, two cases of induced seismicity (2016 Fairview, Oklahoma USA and 2017 Pohang, South Korea) and multiple cases of natural seismicity (including the 2024 Noto Peninsula Earthquake, Japan), were considered. Replication of triggering mechanisms, source properties and time to rupture suggested that stress temporal relaxation and triggered anomalies (STRATA) encompass fundamental hydromechanical processes in seismogenesis. Setting and scale invariance of STRATA suggest it might be a general theory of earthquake nucleation. Based on identified preparatory processes/retrospective validations, a physics-based earthquake forecasting/prediction scheme was proposed. Nurseries/hypocenters of impending earthquakes are identified through simultaneous consideration of locally pre-existing fluid overpressure and spatiotemporal analysis of stress relaxation. Event size and rupture timing are estimated with derived relations herein.