To address the “silent threat” of the iron-based metal corrosion, the development of the precise real-time monitoring and early warning technologies has become a pressing need at present. In this study, a magnetic-responsive liquid-solid triboelectric nanogenerator (MLS-TENG), based on the triboelectrification effect, was proposed to realize mobile and non-invasive corrosion monitoring in iron-based metal environments. By optimizing the device structure and detection system, the MLS-TENG achieved an open-circuit voltage of about 65.2 V and a short-circuit current of about 9.2 μA. Based on this, the “corrosion-resistance-electrical signal” coupling model was established for the first time in this work. By converting the corrosion information into measurable electrical signals, the corrosion status of the iron-based metals in critical components could be effectively distinguished. Inspired by the locomotion of inchworms, a magnetically responsive soft robot was integrated with the liquid-solid triboelectric nanogenerator (LS-TENG), providing a feasible solution for the self-powered corrosion detection across multiple liquid levels under magnetic actuation. Finally, a remote self-powered unattended corrosion monitoring system was established by integrating the STM32 microcontroller and wireless communication module into the MLS-TENG. This self-powered sensing system offers a new solution for long-term health monitoring of the critical infrastructure in the future.