Given the scarcity of experimentally confirmed magnetic structures, the reliable prediction of magnetic ground states is crucial yet remains a long-sought challenge due to the complex magnetic potential energy landscape. Here, we propose a symmetry-guided framework that systematically generates realistic magnetic configurations without requiring any experimental input or prior assumptions such as propagation vectors. The framework incorporates the recently developed oriented spin space group formalism, which captures symmetry-breaking pathways induced by both magnetic ordering and spin-orbit coupling. By performing nonrelativistic and relativistic first-principles calculations, we further establish the energy ladder of the generated magnetic configurations. Exemplified by three prominent unconventional magnets MnTe, Mn${}_3$Sn and CoNb${}_3$S${}_6$, we demonstrate that only a few dozen first-principles calculations are sufficient to identify the ground-state magnetic structure. To demonstrate the universality and robustness of our approach, we conduct large-scale benchmark tests on the MAGNDATA database. Our framework successfully reproduces experimentally reported magnetic structures for 73% of the surveyed materials. Furthermore, in a high-throughput first-principles benchmark involving 305 compounds, 82% of experimentally reported magnetic structures are accurately captured within an energy tolerance of 5 meV per magnetic atom from the computed ground state. Beyond reproducing known magnetic configurations, our framework further predicts a variety of low-energy metastable phases, including altermagnets, spin-orbit magnets, and noncollinear antiferromagnets with spin splitting or geometric Hall effect. Our approach is implemented as the open online program SPINGEN, establishing a general and efficient route toward large-scale prediction of magnetic structures and unconventional magnetic phases.