Threat Modelling

Ask what could go wrong while changes are still cheap to make.

Already have an IT provider? Here's how we work together.

A software company adding online payments

The refund button

The challenge. A small software company is adding a customer portal with payments and refunds. The developers are careful and the code is clean. Nobody has asked what happens if a support account is taken over, or if a customer changes an invoice number in a web address.

The insight. Most serious flaws are in the design, not in a line of code. Drawing the system as a diagram and asking four questions of each part (what are we building, what can go wrong, what will we do about it, did we do it well) finds them while a change is an hour’s work.

The decision. The company made a one-hour threat model a step in its normal design review for any change that touches money or personal data.

The action. We ran the first session with the developers and the product owner. We drew the data flows, listed the threats by part, ranked them by likelihood and effect and picked the few that mattered: a role check on every refund, a limit on refund size without a second approval, and a log entry for every change to a payment record. The developers built them, and we reviewed the result.

The result. The three safeguards shipped with the feature. The company now holds a one-page model for the portal and updates it when the portal changes.

What Security Medic does

Modelling sessions

A working session with the people who build and run the system, producing a diagram, a ranked list of threats and a short list of safeguards.

Models for processes, not only software

The same method for a payment process, a records workflow, a facility or a home network.

Safeguards into the plan

Each chosen safeguard written as a requirement with an owner, so it is built and checked.

Models that stay current

A short update whenever the system changes, so the model never becomes a shelf document.

Where this sits in the cycle

Every client goes through the same six steps, from a town office to a family at home. The Security Medic method

  1. 1. Govern
  2. 2. Design
  3. 3. Assess
  4. 4. Review
  5. 5. Conform
  6. 6. Improve
  • Design. The threat model is part of the design, and shapes the requirements.
  • Assess. The model gives the test plan and the assessment a ranked place to start.
  • Review. Threats ranked by likelihood and effect, reported plainly.
  • Conform. Builders add the safeguards; we check them.
  • Improve. The model is updated with each change and each incident or near miss.

The standards we draw on

The Identify function of the NIST Cybersecurity Framework 2.0, NIST SP 800-30 for risk assessment, the OWASP Application Security Verification Standard and OWASP Top 10 lists for software, and NIST SP 800-82 for operational technology.

Architecture and solution design →Security testing →Incident readiness and recovery →

Getting ahead of it

New features and new AI tools change what can go wrong. Add a short threat check to the steps for approving either.

Start with a free 30-minute call

Tell us what you are building or changing. We will talk through what a first session would cover.

Book a free 30-minute call