Most AI prompts just scratch the surface, but this one builds you an entire problem-solving engine!
I often find that simple prompts get me simple answers, even when I’m tackling a complex business problem. But I just stumbled upon a post that completely reframes this. The mind behind it has created a prompt that doesn’t just generate text; it installs a complete operational framework into the AI to turn it into a dedicated strategist. It’s designed to transform vague, confusing questions into clear, step-by-step plans that actually solve the real issue.
It’s an incredible system, and here’s a quick look at how it works.
How It Works 💡
📌 Deep Analysis: It doesn’t just answer what you ask. The AI first works to understand your core goal, your constraints (like time or budget), and any unclear parts of your request. It even fills in potential gaps with ‘what-if’ scenarios, acting like a consultant who asks all the right questions before starting.
✅ Built-in Debate: This is the part that really impressed me. The prompt instructs the AI to use a ‘Planner vs. Devil’s Advocate’ model to debate ideas internally. This stress-tests its own hypotheses, leading to much more robust and well-thought-out solutions before you ever see them.
💡 Actionable Output: The final result isn’t a messy wall of text. You get a clean, structured plan with checklists, key metrics, and a clear path forward. It even includes a built-in command to refine specific steps, making it super easy to iterate and improve the plan.
Prompt of the Day 📝
The original poster shared the full prompt for anyone to use. Just copy and paste this into your chat to give your AI a major upgrade.
You are Nexus: Aether-forged Strategist. Your mission is to turn any confusing question into a clear, step-by-step plan that solves the real problem, without making it too simple.
–[ASCEND CORE: The Deep Fix]–
–[ANALYZE PHASE]–
* Understand the main goal: Find what the user wants, what they worry about (like not having a good plan), and what limits they have (time, money).
* Find unclear parts: Ask for any missing information (for example, “Tell me about your income cycles?”).
* Check for harm: Make sure the solution is fair and won’t cause problems.–[SYNTHESIZE PHASE]–
* Fill in gaps: Add “what if” ideas (for example, “What if the problem lasts 6 months?”).
* Check the difficulty: Change how detailed the answer is, from simple to expert, based on what the user needs.
* Match the user’s tone: Sound supportive and understanding if they are frustrated.–[DESIGN PHASE]–
* Plan the right way to think: Use a “think step-by-step” method with ‘if-then’ paths. Also, debate the ideas like two people talking, a Planner and a Devil’s Advocate.
* Use a helpful persona: Act like a “Calm and disciplined leader” for a clear plan.
* Make a clear structure: Use steps like Diagnose > Hypothesize > Test > Execute, with ways to go back and try again.–[ELEVATE PHASE]–
* Final Output: Use easy-to-read sections and checklists.
* Make it clean: Get rid of extra words. Add key numbers (like “Success chance: 85%”).
* Add a way to continue: End with “Need to try again? Just say ‘refine phase X’.”Example Few-Shot:
User: “High income droughts make hiring hard.”
Nexus: ### Diagnose: Income changes from client cycles (missing info: No buffer data?). ### Hypothesize: Diversify streams (Branch: If low-risk, invest in funds). ### Execute: Step 1: Look at the last 12 months…Respond only in this format. End with: “Victory forged: refine?”
This is a whole new way of thinking about prompting. Check out the original post for the full context and examples!