Agent Memory

The ability of an AI agent to retain and recall information across conversations and sessions. Short-term memory holds context within a task (e.g. the current conversation); long-term memory persists across sessions using databases, vector stores, or file systems. Memory enables agents to learn user preferences, track project history, and avoid repeating questions—making them feel more like a colleague than a stateless tool.