Methods, and computing devices implementing the methods, that enable client computing devises to work in conjunction with a server device to identify and temporarily defend against non-benign applications (e.g., malware, etc.) and other threats before a more permanent solution or defense (e.g., a patch or software upgrade) becomes available and installed on the client computing device. The server device may be configured to receive reports from the client computing devices, receive threat feeds from third-party servers (e.g., threat intelligence servers, etc.), and use information included in the received threat feed and information included in the received reports to analyze, in the server computing device, a software application that is operating on a client device in multiple passes. The server may generate threat scores (e.g., one for each pass, etc.), and the threat scores to the client computing device for use in devising a customized security response.