How is the system protected against malicious code?

Any malicious code in an external app would only be able to read data that it was given access to - as such these are sandboxed to prevent malicious code being able to harm the client network