The security mindset and threat modeling

Think like an attacker before you write code: map assets, trust boundaries and entry points.

Beginner25 min

Learning objectives

  • Explain the CIA triad and how it applies to web apps
  • Identify trust boundaries in a request flow
  • Run a simple STRIDE threat model on a feature

Why this matters

Security is not a feature you add at the end. Most serious web vulnerabilities come from an assumption nobody questioned:

  • “The client will only send valid data.”
  • “Only logged-in users can reach this route.”
  • “This internal API is safe because nothing outside calls it.”

Each of those assumptions is a bug waiting to be found.

The CIA triad

Every security decision protects one or more of these properties:

  • Confidentiality: only authorised people can read the data.
  • Integrity: data is changed only by authorised people, and only in expected ways.
  • Availability: legitimate users can keep using the system.

Trust boundaries

A trust boundary is any point where data moves into a place with a different level of trust. Everything that crosses a boundary is untrusted input until it has been validated.

[Browser] --HTTP request--> [Web server] --SQL query--> [Database]
    ^ untrusted          ^ boundary: validate      ^ boundary: use
                           every input here          safe queries

Ask of every boundary: who controls this data, and what happens if they lie?

Threat modeling with STRIDE

STRIDE is a checklist of six threat categories. For each feature, ask whether each one applies:

Threat Meaning Typical control
Spoofing Pretending to be someone else Strong authentication
Tampering Changing data in transit or at rest Integrity checks, server-side validation
Repudiation Denying an action happened Audit logs
Information disclosure Leaking data to the wrong party Access control, encryption, careful errors
Denial of service Making the service unavailable Rate limits, quotas, timeouts
Elevation of privilege Gaining rights you should not have Least privilege, authorisation checks

Exercise

Take one feature, such as “reset password by email”:

  1. Draw the data flow from the user to your database and email provider.
  2. Mark every trust boundary.
  3. List at least one threat per STRIDE category.
  4. For each threat, write the control that addresses it.

Key takeaways

  • Treat every input as hostile until it is validated at the boundary.
  • Threat model before you build. It is far cheaper than patching later.
  • Use the OWASP Top 10 as a checklist, not as a complete list of what can go wrong.