What is prompt chaining?

• A. Breaks a complex task into sequential prompts whose outputs feed into the next.
• B. Solves a task in a single prompt without intermediate reasoning.
• C. Generates random prompts to test model robustness.
• D. Retrieves external knowledge without prompting.

Answer: (A)

Prompt chaining means breaking a complex task into a series of prompts where each step's output becomes the input for the next. This creates a pipeline of reasoning, allowing you to handle multi-step tasks by tackling smaller subproblems in order, refining results at each stage, and building up to a final answer. It’s especially useful for planning, data transformation, and iterative improvement, because intermediate outputs can be checked, corrected, or reused before moving on.

Solving in a single prompt skips that step-by-step breakdown, so it can struggle with complexity. Generating random prompts doesn’t establish a deliberate sequence, and retrieving external knowledge without prompting isn’t about chaining prompts at all.