Wireless Sensor Networks (WSNs) are critical to IoT applications but remain highly vulnerable to wormhole attacks, which create covert tunnels between colluding attackers. This research proposes a hybrid intrusion detection framework combining K-Means clustering and Decision Tree classification for efficient and interpretable wormhole attack detection. Experimental evaluation on the publicly available “wormhole attack-Contr2” dataset (637,863 instances) achieved 98.86% accuracy, 97.12% precision, 98.45% recall, and 97.78% F1-score. Comparative analysis with SVM, KNN, Random Forest, standalone Decision Tree, and standalone K-Means demonstrates superior detection performance and computational efficiency, making the proposed framework suitable for resource-constrained WSN and edge environments.