Climate shocks across Central America are rising, yet policymakers lack timely, subnational data to understand how droughts, storms, and food insecurity drive migration. MLEED (Machine Learning for Environmental Event Detection) fills this critical data gap, helping policymakers to link climate impacts with mobility trends and design proactive, evidence-driven adaptation policies. Using new AI-driven data from 25 million news articles, this study maps climate adaptation and disaster impacts across Central America, revealing how drought, storms, and rising food prices shape when, and whether, people migrate toward the U.S.