We study a Bayesian learning model with heterogeneity in which the agents control both short-term manipulations and long-term efforts aimed at solving the optimal contract in dynamic environments. At the same time, considering the dynamic characteristics of contracts, the paper considers the...