The increasing sophistication of cyber threats targeting critical digital infrastructures demands innovative approaches to detection and mitigation. A novel framework, Dynamic Anomaly Pattern Reconstruction (DAPR), was developed to address the limitations of traditional and AI-based methods in identifying and responding to malicious activities. Leveraging adaptive mechanisms and advanced feature extraction techniques, DAPR demonstrated a detection accuracy of 97.5%, outperforming existing methodologies in precision and robustness. Its scalability and efficient resource utilization make it suitable for deployment in varied operational contexts, while its ability to detect previously unseen variants highlights its adaptability to evolving threats. Through a combination of quantitative performance metrics and comparative analysis, the study establishes the framework's significant contributions to the field of cybersecurity. The findings emphasize the importance of real-time, adaptive detection systems in mitigating risks posed by increasingly dynamic and evasive malicious activities.