The rapid growth of the Internet of Things (IoT) in Low-Power and Lossy Networks (LLNs) has highly boosted the possibility of routing-based security attacks, and particularly sinkhole attacks. These attacks, specifically in hop-by-hop models of communication, comprise malicious nodes faking the optimal routes in order to draw in and manipulate traffic, causing loss of data, unauthorized use, and lower dependability. In this paper, we suggest a light and secure mitigation framework that integrates a reputation-based trust model together with private blockchain technology for mitigating metric manipulation-based sinkhole attacks in RPL-based IoT systems. Our approach constantly observes node activities, changes trust scores dynamically, and stores them on an immutable blockchain ledger. Nodes that drop below a preconfigured trust threshold are automatically disconnected from routing activities, thus stopping the propagation of forged routing information. The system is deployed using Contiki OS and the Cooja simulator and tested under normal, attack, and mitigation conditions. Experimental assessment using performance metrics such as packet loss, average routing metric, temperature data accuracy, power consumption, and radio duty cycle demonstrates that the proposed scheme effectively avoids hop-by-hop sinkhole attacks at minimal overhead. The research enhances network robustness, ensures credible data delivery, and enables secure operation in IoT networks with limited resources.