To create reusable prompts, I focus on several key principles. First, I aim for clarity and conciseness, ensuring the core instruction is easily understood and broadly applicable. Second, I implement placeholders or variables (e.g., `[TOPIC]`, `[AUDIENCE]`) where specific information will be inserted later, making the prompt adaptable without rewriting it. I also define the desired output format and structure upfront, such as JSON or bullet points, to ensure consistent results. Providing few-shot examples within the prompt can further guide the model and enhance its reusability across similar tasks. Finally, I advocate for iterative testing and refinement, continuously improving the prompt's effectiveness and versatility based on various use cases. More details: http://energyinnovation.us