Cross-Site Request Forgery (CSRF)
A browser can attach ambient credentials to a request initiated by another site; the server mistakes possession of a cookie for user intent.
Frame the problem
Security starts with a concrete asset, attacker capability and trust crossing.
Why the system fails
A state-changing endpoint accepts a request whose cookies were attached automatically without verifying that the application intentionally initiated it.
The important question is not “what is Cross-Site Request Forgery (CSRF)?” but “which assumption let untrusted data or an over-scoped identity cross attacker-controlled origin → credentialed application request?” Trace the decision at the boundary, then constrain what can happen after the first control fails.
Design the control in layers
Start with the control closest to the interpretation or privilege boundary: Use SameSite cookies plus an unpredictable CSRF token Then add a control that reduces blast radius and telemetry that proves the decision was enforced.
The resulting design is not labelled secure. Record the identified controls, the known failure paths, the remaining exposure, and the evidence you would need during an incident.
| Prevent | Detect | Recover |
|---|---|---|
| Use SameSite cookies plus an unpredictable CSRF token · Validate Origin where appropriate · Never mutate state through safe HTTP methods | Origin mismatches and invalid-token events · Unusual state changes without nearby application navigation | Contain the affected identity or component, scope impact from audit evidence, and preserve a regression test. |
Key points
- Asset: Integrity of state-changing actions performed in an authenticated browser session.
- Boundary: Attacker-controlled origin → credentialed application request
- Primary control: Use SameSite cookies plus an unpredictable CSRF token
- Detection signal: Origin mismatches and invalid-token events
- Always ask what limits damage when the primary control fails.
Boundary control exercise
This lesson uses the shared boundary-control exercise.
Follow the attack
Safe conceptual simulation: capability → missing control → crossed boundary → asset impact.
- 1Attacker starts with: A site that can cause the victim browser to send a cross-origin request.
- 2A state-changing endpoint accepts a request whose cookies were attached automatically without verifying that the application intentionally initiated it.
- 3The weak or missing boundary control is crossed: Attacker-controlled origin → credentialed application request
- 4Impact: Unwanted transfers, configuration changes or account operations as the victim.
- Unwanted transfers, configuration changes or account operations as the victim.
Defend, detect, recover
One prevention is a single point of security failure. Layer it and make failure observable.
- • Use SameSite cookies plus an unpredictable CSRF token
- • Validate Origin where appropriate
- • Never mutate state through safe HTTP methods
- • Origin mismatches and invalid-token events
- • Unusual state changes without nearby application navigation
- • Contain the affected identity or component.
- • Scope access from audit evidence.
- • Fix the boundary and add a regression test.
- • Misconfiguration and new access paths can bypass the intended control.
- • A privileged insider or compromised control plane may still reach the asset.