Think Before You Prompt

The most common mistake

Here’s how most people use AI tools for the first time:

“Make me a website.”

The AI produces something. It’s not what they wanted. They say “no, not like that” and get something else that’s also wrong. After 20 minutes of back-and-forth, they’re frustrated and the project is a mess.

The problem isn’t the AI. The problem is the prompt — and specifically, the thinking (or lack of it) that happened before the prompt.

The 60-second planning exercise

Before you open your CLI tool, spend 60 seconds answering these four questions:

1. What am I building?

Be specific. Not “a website” but “a portfolio site for a freelance photographer with a grid gallery, about page, and contact form.”

2. Who is it for?

This shapes everything. A tool for yourself can be rough. A tool for clients needs to be polished. A tool for beginners needs to be simple.

3. What does “done” look like?

Describe the end result. “When I open the page, I see a dark-themed grid of photos. Clicking a photo opens it full-screen. There’s a navigation bar with three links. The contact form sends an email.”

4. What are the pieces?

Break it into parts. “I need: (1) a navigation bar, (2) a photo grid, (3) a lightbox for full-screen viewing, (4) an about page, (5) a contact form.”

Why this works

AI tools are literal. They do exactly what you ask — which means vague prompts produce vague results, and specific prompts produce specific results.

Compare:

Vague prompt (bad):

Make me a website for my business.

Specific prompt (good):

Create a website for a dog walking business called "Happy Paws" with:
- Homepage: hero image area, three service cards (Daily Walks,
  Pet Sitting, Group Play), testimonials section, contact CTA
- About page: team bios with photos, company story
- Pricing page: three-tier pricing table
- Contact page: form with name, email, phone, message, service dropdown
- Dark green and cream color scheme
- Mobile responsive
- Professional but friendly tone

The second prompt will produce something 10x closer to what you want — and it only took an extra 60 seconds to think through.

The planning template

Here’s a template you can copy and fill in every time:

I'm building: [type of project]
For: [audience]
It needs:
- [feature 1]
- [feature 2]
- [feature 3]
- [feature n]
Style: [visual description]
Tech preference: [framework/language or "your choice"]
When it's done, I should be able to: [describe the end result]

Example filled in:

I'm building: A recipe collection app
For: Myself (personal use)
It needs:
- Add recipes with title, ingredients, instructions, and photo URL
- Browse all recipes in a grid
- Search recipes by name or ingredient
- Tag recipes (dinner, dessert, quick, etc.)
- Filter by tag
- Save to localStorage
Style: Dark theme, warm colors (amber accents), food photography aesthetic
Tech preference: React with Vite
When it's done, I should be able to: Add my favorite recipes, find them
by searching or filtering tags, and see them in a nice visual grid.

Paste that directly into Claude Code or Gemini CLI and watch the magic happen.

The 80/20 of planning

You don’t need to plan every detail. You need to plan the what and the why, not the how.

Plan these (what/why):

• What pages/sections does it have?
• What can the user do?
• What should it look like (general vibe)?
• What constraints matter?

Don’t plan these (how — let the AI decide):

• Which CSS properties to use
• How to structure the JavaScript
• What library handles the routing
• How state management should work

Your plan is the blueprint. The AI is the builder. A good blueprint doesn’t specify what brand of nails to use — it specifies where the walls go and how many rooms to build.