Artificial Intelligence (AI) Library Guide
Frameworks for Prompt Engineering
The CARE Framework
The CARE Framework outlines four core concepts:
| C: Context | Outline the situation—who, what, when, where, why, and how. |
| A: Ask | State what is needed, including any processes or formats to follow. |
| R: Rules | Define constraints such as style, tone, or length. |
| E: Examples | Provide examples (both good and bad) to illustrate expectations. |
Bibliography
Moran, K. (2024, May 24). CARE: Structure for Crafting AI Prompts. Nielsen Norman Group. https://www.nngroup.com/articles/careful-prompts/
The CLEAR Framework
The CLEAR Framework outlines five core concepts:
| C: Concise | Give only essential information. Clear, brief prompts yield better results. |
| L: Logical |
|
| E: Explicit |
|
| A: Adaptive | Revise and refine prompts as needed. Effective prompting is iterative. |
| R: Reflective |
|
Bibliography
Lo, L. S. (2023). The CLEAR path: A framework for enhancing information literacy through prompt engineering. Journal of Academic Librarianship, 49(4). https://doi.org/10.1016/j.acalib.2023.102720
The CREATE Framework
The CREATE Framework outlines six core concepts:
| C: Character |
Indicate who is requesting the information to improve AI tool accuracy. |
| R: Research | Understand the topic well to formulate an effective AI prompt. |
| E: Example | Provide examples to direct the AI toward appropriate responses. |
| A: Audience |
Identify your target audience and determine the information you need to communicate. |
| T: Tone |
Ensure you specify the desired tone of the response, whether professional, casual, humorous, or serious. |
| E: Evaluate | Review each response for accuracy and refine your prompt by incorporating additional guidance or information, if necessary. |
Bibliography
Birss, D. (2023). How to research and write using generative AI tools. LinkedIn Learning. https://davebirss.com/documents/the_prompt_guide.pdf#page=4.07
The ReAct Framework
The ReAct Framework outlines four core concepts:
| Re : Read |
Provide clear instructions, including only what is essential to achieve the desired outcome. |
| A: Answer |
Use a structured approach to ensure the AI accurately interprets context and conceptual relationships. |
| C: Cite | Clearly specify the expected response, addressing its scope, content, format, and complexity level. |
| T: Think | Be ready to rephrase or elaborate your prompts as necessary, recognizing that achieving the desired outcome typically involves an iterative process of experimentation and revision. |
Bibliography
Yao, S., Zhao, J., Yu, D., Du, N., Shafran, I., Narasimhan, K., & Cao, Y. (n.d.). REACT : Synergizing reasoning and acting in language models. Retrieved July 15, 2025, from https://react-lm.github.io/.