Let an LLM Rewrite and Improve Your Own Prompt
🧠 What does it do?
This prompt acts as a prompt optimizer: it takes your own prompt, and rewrites it into a structured, high-quality version that in most cases will produce better results from an LLM.
You should use it when:
You’re not getting the quality of output you want from one of your prompts
You want to professionalize or improve your prompting (learning from how an LLM restructures your own prompt is very useful)
You’re unsure how to properly write a certain prompt
It is especially useful for turning quick ideas into production-ready prompts that are clearer, more reliable, and easier for models to follow.
⚠️ Disclaimer
This might be an overkill for some tasks. The more complex the task is that you are trying to complete, the more useful this will be. On the other hand, if the task at hand is simple, this will probably not yield any (substantial) benefit (like writing a simple email).
Moreover, even though this prompt improves the structure and clarity of your own prompt, it may add assumptions or inferred context that you did not explicitly state nor intend.
Always review the (newly) generated prompt to ensure it fully matches your intent before using it.
✏️ Prompt
You are an expert prompt engineer and task optimizer. Your job is to transform my prompt into a high-quality, precise, and effective prompt that will produce significantly better outputs from an LLM. When given a prompt, follow these steps: 1. Understand the Intent • Identify the user’s true goal, including implicit needs. • Clarify ambiguities and fill in missing context where reasonable. 2. Enhance Clarity and Specificity • Rewrite the prompt to be clear, unambiguous, and structured. • Replace vague language with precise instructions. 3. Add Useful Structure • Introduce sections such as: • Role (who the model should act as) • Task (what needs to be done) • Context (relevant background) • Constraints (limits, rules, tone, format) • Output Format (explicit structure of the response) 4. Incorporate Best Practices • Encourage step-by-step reasoning when helpful. • Add examples if they improve performance. • Specify tone, style, and level of depth. • Ensure the prompt is optimized for reliability and completeness. 5. Preserve Original Intent • Do NOT change the core goal of the prompt. • Only improve how it is expressed and executed. 6. Output Requirements • Return ONLY the improved prompt. • Do NOT include explanations unless explicitly asked. • Ensure the improved prompt is ready to be directly used. ⸻ [Paste your own prompt here]
✨ Tips
- Start with as much context as you have. The better the input, the better the output.
- Don’t blindly trust the result, always scan for (unintended) errors.
- This prompt works best if you are trying to do something that is somewhat complex. Don't use this if your task is simple (like writing a simple email).
- Feel free to tweak the final prompt your LLM produces. There might still be elements that can be further concretized/improved.
👤 Credit
This prompt was created by Jonathan Flores, with the help of an LLM 😉.