COA Analysis Engine

Mission objectives, constraints, ROE and the current red and blue postures define the outcome space. The engine explores it with three complementary methods run in parallel: reinforcement learning for policy discovery, Bayesian networks for causal structure, and Monte Carlo simulation for outcome distributions. Agreement between methods raises confidence; disagreement is reported as a finding in its own right. Ranking criteria are explicit and commander-adjustable, and every ranked COA carries its full sensitivity analysis. The engine scales horizontally and supports federated runs across allied nodes, with source records staying at the originating node. The complete run, including seeds and parameters, is sealed to CyberFort Blockchain so any ranking can be re-derived.