A pre-OS security agent runs in an environment independent of the operating system (OS) but interfaced with the file system and able to exchange information with a security application running over the OS. Prior to the start-up of the OS, an indication of a state or condition is obtained relating to a risk of an inability of the security application to function normally, or to a change in the computer system affecting the start-up of the OS. Based on the indication, a set of one or more actions are determined for resolving the state or condition. The pre-OS security agent executes the set of one or more actions in response to the indication.