Which method splits a complex task into a sequence of smaller prompts that are sent to a large language model (LLM) one after another?

Prompt chaining is the correct answer because it involves breaking down a complex problem into a series of simpler, sequential prompts. The output of one prompt serves as the input for the next, allowing the LLM to build towards a final solution step-by-step. This method enhances the model's ability to handle intricate tasks by managing cognitive load. One-shot prompting provides the LLM with a single example to guide its response, which is different from a sequence of prompts. Tree of thoughts is a more advanced prompting technique that explores multiple reasoning paths and self-evaluates them, rather than simply chaining prompts. Retrieval Augmented Generation (RAG) combines an LLM with an external knowledge base to retrieve relevant information before generating a response; it's about information retrieval, not sequential task decomposition.