Provided are systems, methods, and computer-readable medium for identifying security risks in applications executing in a cloud environment. In various implementations, a security monitoring and management system can obtain application data from a service provider system. The application data can include a record of actions performed by an application during use of the application by users associated with a tenant. The application executes in a service platform provided for the tenant by the service provider system. In various implementations, the application data is analyzed to identify an event associated with a security risk, where the event is identified from one or more actions performed by the application. The system can determine an action to perform in response to identifying the event. In various examples, an agent executing on the service platform can add instrumentation codes used by the application, where the instrumentation provides the application data.