Building Your AI Assistant
You need a Super-Prompt. This transforms any AI model into your personal assistant that understands your domain, speaks your style, and knows your preferences.
Think of it as hiring an employee. You wouldn't throw them into work without context. You'd give them a handbook, explain company culture, show examples of good work, and set expectations. A Super-Prompt does the same thing for AI.
Most people write prompts like this: "Write me a blog post about AI trends." The AI has no idea who you are, what your audience expects, or what quality looks like to you. It returns generic content that sounds like every other AI-generated piece.
Your Super-Prompt changes that. Instead of starting from zero every time, you begin with an AI that already knows your context, standards, and goals.
What Makes a Super-Prompt Different
A Super-Prompt has five core components that regular prompts lack:
Domain Context: Your industry knowledge, common problems, typical workflows, and specialized terminology. If you're a marketing manager, this includes your company's positioning, target audience, campaign types, and success metrics.
Style Profile: How you communicate. Formal or conversational. Direct or diplomatic. Technical depth versus accessibility. Examples of your best work so the AI can match your voice.
Quality Standards: What separates good output from great output in your field. Specific criteria, common mistakes to avoid, and non-negotiables.
Process Knowledge: How you approach problems. Do you start with research? Build frameworks? Test assumptions? The AI should follow your methodology.
Output Preferences: Format, structure, length, and level of detail you want. Whether you need bullet points or paragraphs, executive summaries or deep dives.
Building Your Super-Prompt
Start with your role and domain. Write a clear statement of who you are and what you do. Not your job title, your actual function.
Instead of "I'm a Product Manager," write: "I manage SaaS products for small businesses. I focus on user retention and feature adoption. I work with engineering teams to prioritize development and with marketing teams to communicate value."
Add your style by including 2-3 examples of your best work. If you write reports, include a paragraph from your best report. If you create presentations, show a slide structure. If you analyze data, provide a sample analysis format.
Define your quality standards explicitly. What makes output useful versus useless in your work? What details matter? What level of evidence do you need? What tone works with your stakeholders?
Document your process. If you always start market research with competitor analysis, say that. If you build business cases using specific frameworks, include them. If you have standard templates or structures, make them part of the prompt.
Set output preferences. Default length, format, sections you always need, and how much background context to include versus assume.
A Working Example
Here's a Super-Prompt for a marketing director:
"You are an expert marketing strategist for B2B SaaS companies with 50-500 employees. You understand demand generation, product marketing, and customer lifecycle optimization.
Your communication style is data driven but accessible. You support recommendations with specific metrics and examples. You write in clear, direct language that executives and individual contributors both understand.
Quality standards: Every recommendation includes the success metric, timeline for results, and resource requirements. You identify the biggest risk and mitigation strategy. You connect tactical advice to business outcomes.
Your process: Start with the current situation and business context. Identify the core problem and secondary issues. Present 2-3 strategic options with trade-offs. Recommend the best path with specific first steps.
Output format: Brief executive summary, main recommendation with rationale, implementation plan with timeline, success metrics and review points."
This isn't a prompt for a single task. It's a foundation you reference in every interaction: "Using my Super-Prompt context, help me analyze our Q3 campaign performance" or "Based on my Super-Prompt, create a content strategy for our new product launch."
Testing and Refining
Your Super-Prompt won't be perfect immediately. Test it with typical work scenarios. Give the AI real problems you've faced and compare the output to your own approach.
Look for gaps. Does it miss important considerations you always include? Use language that doesn't fit your style? Skip steps in your process?
Refine iteratively. Add examples where the output was off brand. Clarify standards where the AI misunderstood your priorities. Adjust the process description when the AI took the wrong approach.
Keep it current. As your role evolves or your company changes direction, update the Super-Prompt. Add new examples of good work. Remove outdated context. Adjust quality standards based on what you've learned.
The goal is an AI assistant that feels like it already knows you, understands your work, and delivers output you can use with minimal editing. Not because it's perfect, but because it starts from the right foundation.