Integrated sensing and intelligent interpretation of multidimensional biomarkers are essential for the in-depth characterization of complex biological microenvironments. Here, we developed an integrated AND-gated molecular classifier based on the primer exchange reaction (PER) capable of processing multidimensional biomarkers, including pH, ATP, and miRNA, within a unified molecular computing framework. The system adopts a tandem modular architecture in which chemical inputs are first transduced into single-stranded DNA signals. Together with endogenous miRNA inputs, these signals are specifically recognized by a logic processor via a strand displacement reaction, thereby initiating PER-based primer extension and amplification and ultimately generating label-free, highly sensitive fluorescence decisions. Within this architecture, the system achieves detection limits of 0.18 nM for miRNA and 0.35 μM for ATP and exhibits a precise response across a pH range of 5.0-7.0. Notably, a dual-layer AND-gate configuration generates a strong fluorescence output exclusively under the concurrent presence of low pH, high ATP, and elevated miRNA levels, as demonstrated in both buffer-based and cellular models. This work advances the PER from a nucleic acid amplification tool to a programmable molecular decision-making engine, providing a versatile molecular computing platform for the intelligent diagnosis of complex disease states.