The present paper examines what domain-general causal knowledge reasoners need for at least some outcome-variable types to construct useable content-specific causal knowledge. In particular, it explains why it is essential to have analytic knowledge of causal-invariance integration functions: knowledge for predicting the expected outcome assuming that the empirical knowledge acquired regarding a causal relation holds across the learning context and an application context. The paper reports two experiments that support the hypothesis that preschool children have such knowledge regarding binary causes and effects, enabling them to generalize across contexts rationally, favoring the causal-invariance hypothesis over alternative hypotheses, including interaction (e.g., linear) integration functions, heuristics, and biases.