Designing instructions that get accurate, useful output from AI language models across multi-step workflows.
Prompt engineering is the practice of designing effective inputs for large language models. It spans from writing clear queries to architecting multi-step workflows with chain-of-thought reasoning, system prompts, and agent orchestration, bridging human intent and machine capability.
A 5-level prompt competency framework (Beginner-Expert) defining behavior-based proficiency boundaries, used to calibrate checklist difficulty across levels.
Systematic learning path from basic prompting to agent design, providing criteria for progressive technique difficulty and advanced technique classification.
Systematic learning path covering clear instructions, example provision, XML structuring, role prompting, thought elicitation, and prompt chaining, used as technical evidence for L2-L4 checklist items.
Empirical evidence that chain-of-thought prompting elicits reasoning in large language models, providing academic authority for the L3 CoT checklist item and the overall technique difficulty hierarchy.
