MInD is a Python-based memory framework designed to augment LLM-driven AI agents with robust memory capabilities. It enables agents to capture user inputs and system events as episodic logs, condense those logs into semantic summaries, and retrieve contextually relevant memories on demand. With configurable retention policies, similarity search, and automated summarization, MInD maintains a persistent knowledge base that agents consult during inference. This ensures they recall prior interactions accurately, adapt responses based on history, and deliver personalized, coherent dialogues across multiple sessions.