Phase D — Decision

Select the countermeasure.

Objective

Given a positively identified platform (Phase B) with known vulnerabilities (Phase F pre-engagement intelligence), select the optimal countermeasure or countermeasure chain. Phase D is the bridge between intelligence and action.

Phase D does not execute anything. It produces a ranked decision: given what we know about this target, what is the highest-probability method to achieve the desired effect?

Decision Inputs

Input	Source
Platform identification	Phase B
Tracking data quality	Phase C
Failsafe decision tree	Phase F (pre-engagement) / database
Exploit chain library	Database
Adversary hardening tier	Phase B + Phase F assessment
Environmental conditions	Operational context

The Failsafe Decision Tree

The core of Phase D for most autonomous platforms. Every failsafe is a guaranteed state transition that you can trigger on purpose.

The decision tree inverts the designer’s intent: safety mechanisms become attack surfaces.

IF target has GPS dependency AND GPS is not authenticated:
    → GPS spoofing viable (Mirage techniques)
    
IF target command channel is unencrypted AND unsigned:
    → Command injection viable (Override techniques)

IF target decision thresholds are known AND triggerable:
    → Threshold manipulation viable (Fracture techniques)

IF target has finite consumable resources with known depletion rates:
    → Resource exhaustion viable (Sever techniques)

Countermeasure Ranking

Phase D ranks viable countermeasures by:

Success probability — tested success rate from the database
Time to effect — initiation to platform state change
Operational constraints — legal, environmental, equipment
Reversibility — can the effect be undone if engagement is called off
Detection risk — does the countermeasure reveal capability

Hardening Tier Filtering

The target’s hardening tier filters technique categories before ranking:

Tier	Viable Categories	Filtered Out
0 (COTS unmodified)	All	None
1 (Protocol-hardened)	Specter, Mirage, Fracture, Sever	Override (requires bypass)
2 (Sensor-hardened)	Specter, Fracture, Sever	Override, Mirage (requires defeat of authentication)
3 (Full-stack hardened)	Specter	All active categories require novel research

Phase D must be honest. If no viable countermeasure exists at the assessed hardening tier, the output is “no viable countermeasure with current capability” — not a low-probability guess.

Human Authorization

Phase D produces recommendations. Phase E requires human authorization before execution. The decision engine does not fire autonomously. An operator reviews the ranked list, confirms the engagement, and authorizes execution.

This is a design constraint, not a technical limitation.

Phase D Outputs

Output	Format	Feeds
Ranked countermeasure list	Ordered exploit chain references	Phase E execution
Per-countermeasure success estimate	Probability	Phase E go/no-go
Required equipment and parameters	Configuration data	Phase E setup
Engagement feasibility	Go / conditional / no-go	Operator decision