Agent-based patching is a software update model in which a lightweight endpoint agent receives patch jobs from a central management platform and applies them locally. The agent usually reports device status, checks policy, downloads the update, and can trigger installation or reboot steps on schedule.
This matters in cyber security because unpatched systems are a common entry point for attackers. A good agent-based system can speed deployment, verify that fixes landed, and help defenders prioritize exposed vulnerabilities. It is especially useful in cloud-managed environments and distributed fleets where running a dedicated patch server would add overhead. The tradeoff is dependence on the agent and its communication path: if the agent is disabled, lacks permissions, or cannot reach the control plane, patching can fail and the exposure window stays open.