The rapid adoption of electric vehicles (EVs) is fundamentally transforming the automotive industry, prompting a surge in the installation of charging stations to accommodate the growing number of EVs and enhance overall mobility and user experience. Efforts to conduct machine learning-based cybersecurity research and develop solutions to address the growing threats and vulnerabilities in EV charging station infrastructure face challenges stemming from the unavailability of suitable datasets. The primary contribution of this study is addressing these challenges by publishing a multi-dimensional dataset that comprises power consumption data, network traffic and host activities of the EVSE in both benign and attack conditions. The experimental testbed utilizes a real EVSE, Raspberry Pi and standard industry communication protocols for EV charging infrastructure, with the scenarios observing the EVSE in both idle and charging states. The results of statistical analysis and machine learning classification tasks demonstrate the suitability of this dataset for baseline behavioral profiling, classification and anomaly detection tasks.