I'm talking more of an OS level security protocol... best practice guide. If it is of any help, on all my systems here, I deploy SELinux - a security framework I am extremely happy with. All my system components (packages, programs etc) are confined in their own security realms by deploying appropriate security policies.
Most of these policy were designed from scratch - by myself, other were changed from the "standard" ones provided by my distro to reflect the particular settings of my host environment. I have been using SELinux for years and I can vouch for it - it provides sufficiently good (understand deep) level of security - it goes to the Linux core. As I am fairly new to FS (until about a month ago I didn't even know it existed and initially wanted to use and deploy Asterisk, but soon realised the error in my ways ;-) ), I am in a process of writing one such security policy for FS, which, after the appropriate testing and possible tweaking, will be deployed on the system I intend to use FS on. If there is a sufficient interest I will provide this new policy to everyone interested - as I already pointed out, SELinux provides a very good level of security when deployed properly, so it is capable of confining packages/applications where they belong - in their own security realm. As far as the non-OS level of security goes (i.e. securing FS itself from internal exploits of the protocol(s) it uses, including toll fraud etc) then I won't be of much use until I know more about FS - something I am in a process of correcting as I do read up about how FS works from various sources, including the FS book (and soon to get hold of the cookbook as well).