This document defines a formal, conditional framework for modelling trust under uncertainty. Trust is treated as a retrospective system variable governed by explicit rules for accumulation, decay, loss, and decision thresholds. The framework is non-moral, non-psychological, and non-prescriptive, and makes no claims of optimality or empirical calibration. The framework applies to bounded systems operating under uncertainty, including simulations and agentic systems, and is intended for inspection, critique, or comparison rather than implementation. Version v0.2 is released as a frozen public record.
Gew Kok Pian (Tue,) studied this question.