AI will always give you some response. The problem is that “some response” often feels generic, flat, or slightly off—especially when you’re in a hurry and you don’t want to spend 15 minutes rewriting your prompt.
Here’s the quick hack: instead of guessing how to improve your prompt or using a prompt generator, make the AI grade it and then tell you exactly what’s missing.
The method: “Rate it / Make it a 10”
The approach is ridiculously simple and works with basically any prompt:
- Step 1: Ask the AI to rate your prompt from 1 to 10.
- Step 2: Ask: “What would make it a 10/10?”
That’s it. Two lines. No prompt engineering rabbit hole. No endless tweaking. You’re outsourcing the improvement process to the model itself.
Why this works so well
Most people try to fix prompts by instinct: adding random details, rewriting sentences, hoping the AI “gets it.” But the AI can’t read your mind—so your prompt needs the right ingredients.
When you ask for a rating, the AI is forced to evaluate clarity and completeness. When you ask “what makes it a 10,” it shifts into editor mode: it lists what’s missing (context, audience, tone, constraints, structure, examples, formatting), and often suggests a stronger version of your prompt.
The result: sharper output with less effort.
A real example: the follow-up email that comes out bland
Let’s say you want a follow-up email after a meeting. Your basic prompt might be:
“Write an email summarizing the meeting’s discussion points.”
You’ll get an answer—but it may feel too generic, too polite, too “template.” It doesn’t pop.
Now use the method:
1) Rate the prompt:
“Rate this prompt from 1 to 10: ‘Write an email summarizing the meeting’s discussion points.’”
Whether it says 3/10, 6/10, or even 9/10, you follow with:
2) Force the upgrade:
“What would make it a 10 out of 10?”
Suddenly, the AI tells you what’s missing—like:
- Who you’re emailing (client, manager, teammate)
- The tone (warm, crisp, confident, friendly)
- The goal (confirm next steps, ask for approval, schedule a follow-up call)
- A structure (bullets for decisions, action items, deadlines)
- Any key details to include (names, dates, agenda topics)
And the best part: you can add one more line:
“Now rewrite my prompt as a 10/10.”
Boom. You get a stronger prompt and a better email with minimal work.
The copy-paste version you can use on anything
Use this template for any task—emails, blog outlines, titles, social posts, code, scripts, summaries, anything:
Rate this prompt from 1 to 10:
[PASTE YOUR PROMPT]
What would make it a 10/10?
Rewrite it as a 10/10 prompt.If you want even more control, add one final instruction:
After rewriting it, ask me the 3 most important missing questions before you answer.That’s how you go from “okay output” to “wow, that’s exactly what I meant.”
When to use this method (spoiler: almost always)
This hack shines when:
- Your output feels generic or “corporate bland”
- You don’t know what detail the AI needs
- You want better results without learning prompt engineering
- You’re repeating the same task and want a reusable “perfect prompt”
The takeaway
The fastest way to improve prompts isn’t to rewrite them blindly—it’s to ask the AI to diagnose them.
Next time your AI answer feels off, don’t wrestle with the prompt. Just run:
“Rate it / Make it a 10.”
Try it on your next prompt and see how much cleaner your results get—without the trial-and-error spiral.
