Final installment in “A Guide to Threat Modelling for Developers”
Feedback and continuous improvement is central to managing risk.
Neither the systems we build nor the threats they face are simple, as I
stressed at the start of this guide. And every team is different- with
different skills, tools, constraints and personalities. There is no single
way to threat model, this guide simply provides some basics to get you
started. Much like test-driven development or continuous delivery, threat
modelling rewards investment.
One way to improve is to perform a retrospective on your threat
modelling efforts, once you have run a few sessions. Ask what went well
and what could be improved. Is the timing right? Was the scope too
granular? Not granular enough? What about the location or remote tools you
have used? What issues cropped up after the session? How long did the
scope take to deliver? By asking such questions, the team will adapt and
build mastery over time, doubling down on what works and discarding what
adds little value.