Task Chaining

A concept in AI assistant architecture referring to the serial execution of multiple independent capability modules, where the output of each step automatically becomes the input of the next. It contrasts with early single-turn question-answering models and is commonly implemented using Function Calling and multimodal models.