risk-diagnostics

An early-warning system that models disasters as instability transitions rather than isolated events. It combines force-based instability modeling with an interpretable ML escalation-risk layer to detect when hazards become disasters due to exposure growth, response delays, and buffer collapse.

Block-level diagnostics for stress-testing Ethereum fee behavior under EIP-1559.