What if the fastest way to improve your AI prompts was to ask the AI to do it for you?
By Linda Pophal, MA, SPHR · Strategic Communications, LLC
Prompt engineering is the practice of crafting precise, well-structured instructions for generative AI tools—specifying context, desired format, tone, audience, and output type—to produce more accurate, relevant, and useful responses. As AI tools become standard in professional workflows across every industry, the ability to write effective prompts is rapidly emerging as one of the most consequential skills in the modern workplace. And like most skills, it improves with deliberate practice not just use.
I’ve been an early adopter of AI tools. I’ve used tools like ChatGPT, Claude, and Perplexity extensively in my own work and for clients, and I’ve been intentional about trying to improve my prompt writing skills along the way.
More specific context. Clearer output instructions. Better framing of what I actually need.
Recently I had an epiphany—a genuinely simple idea that I’m a little embarrassed I didn’t arrive at sooner.
Instead of trying to figure out how to write better prompts on my own, I could just ask the AI to help me improve them!
The self-improving prompt strategy
Here’s the approach. Before submitting any prompt, I add one additional line at the end:
Add this line to the end of any prompt:
“Before proceeding, please improve this prompt to make your response better and as actionable as possible. Also provide me with a list of tips that I should incorporate into the next prompt I write.”
That’s it. Two sentences. But the impact is significant.
The AI does two things in response. First, it rewrites your original prompt—often in ways that are noticeably more specific, better contextualized, and more likely to produce the output you actually want. Second, it gives you a personalized list of prompt-writing tips based on what was missing or improvable in the prompt you submitted.
You get a better response right now from the improved prompt. And you get a coaching lesson for the next one.
The continuous improvement cycle
What makes this strategy particularly useful is that it compounds. Here’s how the cycle works:
Round 1: Write your prompt. Add the improvement line. The AI improves the prompt, executes it, and gives you tips.
Round 2: Apply those tips to your next prompt. Add the improvement line again. The AI improves the new prompt, executes it, and gives you updated tips.
Round 3 and beyond: Repeat. Each cycle, your baseline prompt quality improves. Each cycle, the AI’s suggested improvements get more refined—because you’ve already incorporated the previous round’s feedback.
Over time, this builds genuine skill. You start to internalize the patterns—what level of context AI tools need, how to specify output format, when to provide examples, how to frame constraints. The AI is essentially functioning as a personal writing coach for one of the most important new skills in the modern workplace.
Building a prompt library
The second part of the strategy is equally practical: as you develop and refine prompts for specific recurring tasks, save them.
A prompt library—even a simple document or folder—gives you a starting point for every type of task you use AI for regularly. Blog post drafts. Email responses. Research summaries. Meeting prep. Social media copy. Rather than starting from scratch each time, you start from a proven prompt that you’ve already refined through multiple improvement cycles.
And because you’ll always be adding that improvement line, the library continues to evolve. The prompt you saved six months ago will be meaningfully better six months from now—because you’ll have run it through the improvement cycle dozens more times and incorporated what you learned.
This is, at its core, the same discipline I’ve written about frequently: continuous improvement grounded in intentional practice. As I noted in my post on what old-school journalism still teaches modern content marketing, the fundamentals of good communication—clarity, specificity, audience awareness—are the same whether you’re writing a news article, a blog post, or an AI prompt. The improvement cycle simply makes that learning faster and more systematic.
A few practical tips to get started
Start with a prompt you use regularly. Pick a task you do frequently with AI—a type of email you write, a content format you produce, a research query you run. Apply the improvement strategy to that prompt first. You’ll see the most immediate value in contexts where you already have a sense of what a good output looks like.
Save both versions. When the AI improves your prompt, save the improved version alongside your original. Over time, comparing the two reveals patterns in what you consistently under-specify—context, format, audience, constraints—that you can start catching on your own.
Don’t skip the tips list. The list of prompt-writing tips the AI generates is often more valuable than the improved prompt itself. Read them. Apply them. They’re tailored to your specific prompt, which makes them far more actionable than generic advice about prompt writing.
Treat the library as a living document. A prompt library that isn’t updated is just a collection of old prompts. Schedule a brief monthly review—which prompts are you using most? Which are producing the best results? Which need another round of improvement? Twenty minutes a month keeps the library genuinely useful rather than ceremonially maintained.
Am I late to the party?
Possibly. This may be something many of you are already doing, in which case I’d genuinely love to hear what results you’ve seen and what other practices you’ve found useful. The AI tools themselves are evolving rapidly—prompt strategies that worked six months ago may be less necessary today, and new approaches emerge constantly.
But if this is new to you, try it today. Write a prompt you were already going to write. Add those two sentences at the end. See what the AI does with it.
I’d be surprised if you don’t immediately see the difference.
Frequently asked questions about AI prompt writing
What is prompt engineering and why does it matter?
Prompt engineering is the practice of crafting precise, well-structured instructions for generative AI tools to produce more accurate, relevant, and useful responses. It matters because the quality of AI output is directly tied to the quality of the input—a vague or poorly structured prompt produces a generic or off-target response, while a specific, well-contextualized prompt produces something genuinely useful. As AI tools become standard in professional workflows, prompt writing is emerging as one of the most consequential skills across virtually every industry.
How do I write better AI prompts?
Effective AI prompts typically include: clear context (who you are, what you’re trying to accomplish, and why); a specific output format (length, structure, tone, audience); relevant constraints (what to include, what to avoid); and examples where helpful. One of the fastest ways to improve your prompts is to add this line at the end of any prompt: ‘Before proceeding, please improve this prompt to make your response better and as actionable as possible. Also provide me with a list of tips that I should incorporate into the next prompt I write.’ The AI will rewrite your prompt and coach you on what to do differently next time.
What should I include in an AI prompt library?
An AI prompt library should include refined, tested prompts for the tasks you perform most frequently with AI tools—content drafts, email templates, research queries, meeting prep, data analysis requests, and any other recurring use cases. Save both your original and the AI-improved versions of each prompt, along with notes on what outputs they produce. Review and update the library monthly. The goal is a living resource that gives you a strong starting point for any task, rather than a static archive of prompts that quickly become outdated.
How can I use AI to improve my AI prompt writing skills?
Add a self-improvement instruction to every prompt you write: ‘Before proceeding, please improve this prompt to make your response better and as actionable as possible. Also provide me with a list of tips that I should incorporate into the next prompt I write.’ The AI will rewrite your prompt (often significantly improving it) and give you personalized coaching for the next one. Applying those tips consistently, prompt after prompt, builds genuine skill over time—turning every AI interaction into a learning opportunity rather than just a transaction.
Are you already using a process like this to improve your AI prompt writing? What results have you seen, and what other best practices have you found most valuable?
