Threat Modelling Foundations, 20 min read

1.R Review and references

In this lesson you will:

  • Recap the main points of Module 1
  • Test yourself on six review questions
  • Find the sources cited in this module

This page closes Module 1, Threat Modelling Foundations. Citation numbers in square brackets in the sections, such as [1], refer to the reference list at the end of this page.

Module summary

  • Threat modelling analyses a representation of a system to find security and privacy concerns early, and records decisions about them.
  • The four questions structure every method: What are we working on? What can go wrong? What are we going to do about it? Did we do a good enough job?
  • Assets are protected; threat actors cause threats; vulnerabilities let threats succeed; risk is likelihood × impact; controls reduce risk.
  • Design flaws such as missing authorisation are invisible to scanners and cheapest to fix at design time.
  • The Threat Modeling Manifesto values people and collaboration, continuous refinement and doing over tools, snapshots and talk. Avoid the hero threat modeller, admiration for the problem and the perfect representation.
  • Data flow diagrams model the system with external entities, processes, data stores, data flows and trust boundaries. Every flow starts or ends at a process.
  • Trust boundaries are where the level of trust changes. Flows that cross them carry most of the serious threats.
  • Work at Level 1; zoom in where a box hides a boundary or a security decision.
  • Record assets, entry points and dated assumptions, and scope the exercise to include where its data is used downstream.

Review questions

Answer all six before opening the answers.

1. Which statement BEST describes the output of a threat modelling exercise?

  • A. A list of exploitable vulnerabilities confirmed in the running application
  • B. A model of the system, a list of threats, decisions with owners and evidence of validation
  • C. A compliance checklist showing which controls from a standard are implemented
  • D. A risk register of organisation-wide risks with annual loss expectancy figures

2. "Settlement files are not signed" is BEST described as:

  • A. A threat
  • B. A vulnerability
  • C. A risk
  • D. A control

3. A DFD shows "Ops analyst" writing directly to "Settlement folder". What should the facilitator do?

  • A. Accept it; staff are trusted inside the network boundary
  • B. Add the process that performs the write, then analyse that process and the flow
  • C. Delete the flow because external entities cannot touch stores
  • D. Replace the store with a process so the diagram is valid

4. Where should a team with only 30 minutes focus its threat analysis?

  • A. On the processes that contain the most lines of code
  • B. On the data flows that cross trust boundaries
  • C. On the data stores, starting with the largest
  • D. On the external entities, starting with the least trusted

5. Which is an example of the "admiration for the problem" anti-pattern?

  • A. One senior architect writes every threat model for the company
  • B. A session spends 90 minutes on clever attack chains and records no decisions
  • C. A team refines the diagram for three weeks before analysing it
  • D. Threat models are filed after release and never updated

6. Why should assumptions be written down with a date and source?

  • A. Auditors require every assumption to be signed by a manager
  • B. A false assumption is a common root cause of serious flaws, and recording it prompts someone to verify it
  • C. Assumptions replace the need to draw trust boundaries
  • D. Dated assumptions make the threat model valid for twelve months
Show answers
  1. B. A threat model contains the system model, threats, decisions and validation. A is a pen-test report; C is a compliance check; D is an enterprise risk assessment.
  2. B. It describes a property of the system that would let a threat (tampering with the file) succeed.
  3. B. Data does not move by itself. The missing process is where access control should happen, so it must be on the diagram and analysed.
  4. B. Boundary crossings are where trust assumptions change and where most serious threats sit.
  5. B. Admiration for the problem means fascination with attacks without reaching decisions. A is the hero threat modeller; C is the perfect representation.
  6. B. Writing assumptions down exposes them to checking. Kestrel's "only the job writes to the folder" assumption was false.

References

  1. Threat Modeling Manifesto working group. Threat Modeling Manifesto. 2020. https://www.threatmodelingmanifesto.org/
  2. Shostack, A. Threat Modeling: Designing for Security. Wiley, 2014.
  3. OWASP Foundation. OWASP Top 10 (2021 and 2025 editions), A01 Broken Access Control. https://owasp.org/Top10/
  4. CERT-In. Directions under sub-section (6) of section 70B of the Information Technology Act, 2000, 28 April 2022 (six-hour incident reporting).
  5. Shostack, A. Threats: What Every Engineer Should Learn from Star Wars. Wiley, 2023.
  6. OWASP Foundation. Threat Modeling Cheat Sheet and Threat Modeling Process. https://cheatsheetseries.owasp.org/cheatsheets/Threat_Modeling_Cheat_Sheet.html