Detection Engineering
Telemetry becomes a detection hypothesis, an alert, an investigation and a response path; a noisy alert with no owner is not a control.
Frame the problem
Security starts with a concrete asset, attacker capability and trust crossing.
Why the system fails
The system logs volume without identity/resource context, alerts on vague anomalies, or has no tested response attached.
The important question is not “what is Detection Engineering?” but “which assumption let untrusted data or an over-scoped identity cross security-relevant event → actionable responder signal?” 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: Design telemetry alongside the control 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 |
|---|---|---|
| Design telemetry alongside the control · Write detections as testable hypotheses with owners · Connect each alert to a containment action | Coverage tests, alert precision and time-to-triage | Contain the affected identity or component, scope impact from audit evidence, and preserve a regression test. |
Key points
- Asset: Time-to-detect and the evidence required to contain compromise.
- Boundary: Security-relevant event → actionable responder signal
- Primary control: Design telemetry alongside the control
- Detection signal: Coverage tests, alert precision and time-to-triage
- 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: An adversary using valid-looking actions after prevention fails.
- 2The system logs volume without identity/resource context, alerts on vague anomalies, or has no tested response attached.
- 3The weak or missing boundary control is crossed: Security-relevant event → actionable responder signal
- 4Impact: Attacker dwell time grows and incident scope remains unknown.
- Attacker dwell time grows and incident scope remains unknown.
Defend, detect, recover
One prevention is a single point of security failure. Layer it and make failure observable.
- • Design telemetry alongside the control
- • Write detections as testable hypotheses with owners
- • Connect each alert to a containment action
- • Coverage tests, alert precision and time-to-triage
- • 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.