API Overview

High-level modules: - ThreatModeling.threat_model_creator.build_threat_model – orchestrates full threat modeling pipeline. - ThreatModeling.tm_graph.tmr_to_graph – converts threat model result dict into a NetworkX DiGraph. - AttackGraphGeneration.attackgraph – builds and explores attack graphs. - TTCComputation.system_ttc / kube_ttc – TTC aggregation and per-component estimators.

Offline mode surfaces through environment variable TC_OFFLINE and short-circuits LLM calls.

Module Map

Area	File	Purpose
Threat modeling orchestration	ThreatModeling/threat_model_creator.py	Stateful pipeline graph
Technique analysis	ThreatModeling/technique_analysis.py	Technique generation / summarization
Asset categorization	ThreatModeling/asset_categorizer.py	Deterministic or LLM-based grouping
System model	ThreatModeling/system_model.py	Graph wrapper helpers & drawing
Attack graph	AttackGraphGeneration/attackgraph.py	Stochastic walks & statistics
TTC core	TTCComputation/kube_ttc.py	Per-node TTC math
TTC aggregation	TTCComputation/system_ttc.py	Hierarchical TTC propagation

Extension Points

• Replace LLM provider: adjust get_deepinfra_model() implementation.
• Add tactic filtering: post-process tactics dict before technique expansion.
• Extend TTC weighting: subclass KUBE_TTC with alternative formulas.