A Review of Prompt Engineering Techniques for Large Language Models

This paper presents a comprehensive review and a conceptual framework for designing and evaluating textual instructions (prompts) for large language models (LLMs). The motivation for this review stems from the increasing importance of prompts as the primary interface for LLMs and the need to overcome prompt brittleness, which undermines reliability and reproducibility. From this lens, this review introduces a structured methodology for the systematic design, testing, and refinement of prompts, including quantifiable methods for assessing efficiency and resolving ambiguity. We also establish guidelines for prompt drafting, experimental testing, version control, and multi-tiered evaluation protocols. Furthermore, this review examines the influence of user learning curves and how organizational contexts influence prompt effectiveness by examining human-AI collaboration dynamics. Our review then concludes by identifying some of the commonly seen challenges and showcasing future research opportunities in this thematic research area.