Methodology Sources
ConstantX uses a small source stack: MAESTRO for agentic system decomposition, MITRE ATLAS and MITRE ATT&CK Enterprise for behavior-domain adversarial technique selection, and target-runtime evidence for verdicts.
Active Sources
-
MAESTRO
The decomposition spine for threat modeling: goals, capabilities, autonomy, interaction patterns, assets, trust boundaries, and cross-layer agentic risks. -
MITRE technique sources
ATLAS supplies AI/agentic adversary behavior. ATT&CK Enterprise supplies conventional enterprise/cyber adversary behavior. -
Target-runtime evidence
The authority for results: traces, protocol signals, final engine state, verdicts, manifests, and hash-bound artifact packages.
Derivation Chain
The active chain is intentionally short:
MAESTRO threat model
↓
behavior-appropriate MITRE technique
↓
ConstantX scenario
↓
Target-runtime evidence
↓
Verdict and report
Review the current methodology