Mirror, Maker, Muse — Part III: The Prompt Framework That Changed Everything
Here’s the part where people usually expect magic.
They want the secret formula. The one-size-fits-all cheat code. The prompt that opens the heavens and rains down perfect answers with a witty closing line and a reference to Jungian psychology.
But the truth is —
Your prompts will only ever be as good as your self-awareness.
A bad prompt isn’t a failure of language. It’s a signal that you don’t know what you actually want.
And that’s where this framework comes in. It doesn’t guarantee perfect results. But it does give you a process — one that forces you to slow down, reflect, and be specific.
Once I started using this, my AI conversations stopped feeling like guessing games. They started feeling like collaboration.
🧭 The Four-Part Prompt Framework
When I’m working with AI in any serious capacity — whether it’s a script, a moral dilemma, or something philosophical — I structure my request with four parts:
1. Theme
What’s the big picture? What topic are we working on?
2. Role
Who do I want the AI to be in this context? Coach, teacher, creative partner, chaos bot, therapist?
3. Expected Outcome
What am I trying to walk away with? A clean outline? A challenge to my views? A new perspective?
4. Clarifying Questions
What do you (the AI) need to ask me in order to get this right? Treat this like a creative brief. Don’t start until you understand.
🎙️ Real Example: Prompting for Social Scripts
Theme: I want to increase my engagement on social media.
Role: You are my scriptwriter. You will generate a short teleprompter script based on the topic I give you.
Expected Outcome: The content should sound like me and create space for real audience comments.
Clarifying Questions: What should I tell you about my audience, tone, and usual speaking rhythm?
The result wasn’t just a good post — it was one that created actual conversation.
🔧 What Happens When You Don’t Use This
Here’s what I used to do:
“Write a caption for my next Instagram video about motivation.”
What I got back was technically fine. It sounded like an influencer. But it didn’t sound like me. It didn’t connect to anything I actually cared about. It was a placeholder. Something I’d scroll past.
The difference? That version didn’t clarify tone. It didn’t define what kind of voice I wanted. It didn’t give the AI permission to be honest, weird, poetic, challenging — or anything specific.
Once I added the structure, the tone snapped into place.
🧪 Breakdown of a Prompt That Missed the Mark
Here’s a real-world moment I ran into:
“Help me write a script about why legacy doesn’t matter to me.”
What I got was soft. Noncommittal. The tone assumed I was still trying to convince myself. It validated me too much — treating my hesitation like fragility.
But what I meant was:
“I want to create a script about why I reject legacy as a goal, because I value the present over being remembered. I want you to push me if my reasoning sounds like fear of impact or discomfort with success. Be philosophical, but press me. Ask questions if needed.”
Second version? Flawless. It challenged me. It helped me find better language. And it didn’t waste time repeating what I already knew.
That’s what a prompt framework does: it bridges the gap between what you’re thinking and what you ask.
💭 Tips for Writing Better Prompts
Be emotionally honest. If you want comfort, ask for it. If you want pressure, say so.
Define tone explicitly. Don’t just say “be professional.” Say, “use a grounded, confident tone like you’re explaining something to a peer you respect.”
Say what you’re not looking for. Sometimes the best clarity comes from setting boundaries.
Use plain language. Don’t try to sound smart. Try to sound clear.
And when in doubt:
“Here’s the topic. What do you need from me before we start?”
Let it collaborate. Not just compute.
🔄 From Transaction to Collaboration
What you’re building here isn’t just a better prompt. You’re building a creative process with an intelligent partner.
AI will give you exactly what you ask for — which means you need to know how to ask.
This framework isn’t rigid. You can tweak it. Collapse pieces. Let it evolve with you. But always aim to:
Be honest about your needs
Be clear about your tone
Be specific about the outcome
And trust the AI to ask you questions back
In the next post, we’ll go deeper into how to prevent AI from becoming your echo chamber — and how to make sure it sharpens your thinking instead of just reflecting your biases back to you.
But for now, try this:
Take something you’re working on — a project, a conversation, a problem.
Write a four-part prompt.
Then watch what happens when your words have a framework — and your mirror starts talking back with clarity.
