Modern cybersecurity threats require advanced tools for analyzing attack scenarios that evolve over time. We use temporal attack graphs to model the progression of evolving threat landscape, since existing tools often generate static attack graphs, which have limited practical utility. This paper proposes a framework to address these limitations and automatically generate temporal attack graph (TAG) datasets using containerized environments and attack graph generation tools. Beyond attack modeling, TAGs offer critical advantages in improving intrusion detection and proactive defense. Our approach implements a time-window based mechanism to simulate the evolution of the network in a realistic environment, capturing the temporal aspects of attack scenarios. By combining infrastructure-as-code tools, vulnerability scanning systems such as Nessus, and reasoning engines such as MulVAL, the framework models how attack surfaces evolve over time paving the way for intelligent and time-aware intrusion detection systems.