The Prompt Pack

The prompts I wish I'd had at prompt one.

70 field-tested prompts for building real apps with AI when you don't know how to code. Each one comes with when to use it, the prompt itself, and why it works. 10 are free below — the other 60 live in the Pack.

The Full Pack — $29

60 more prompts • MVP scoping • debugging • design • backend • launch • one-time payment • instant access

The 10 free prompts

Getting StartedFree

The Master First Build

When to use: The very first prompt for a brand new project. Use this when you have an idea and an empty Lovable project.

Build me a web app called [YOUR APP NAME]. In one sentence, it helps [YOUR USERS] do [WHAT IT DOES]. The homepage should have a clear headline, a short subheadline, one primary call-to-action button that says [BUTTON TEXT], and three sections underneath that explain how it works. Use a clean, modern design that would look at home in 2026. Do NOT add authentication, a database, or any backend yet. Just build the front-end shell so I can see it and react to it.
Why it works

It gives the AI a full picture (what, who, why) but strictly limits scope to the front-end so you don't drown in complexity on day one.

Getting StartedFree

Define My MVP

When to use: Before you build anything, use this to talk the idea through with the AI and pick the smallest version worth shipping.

I have an idea for an app that does [WHAT YOUR APP DOES] for [YOUR USERS]. I''m not technical. I want to ship the smallest useful version first. Ask me 5 clarifying questions to help you understand what really matters. Then, based on my answers, propose an MVP that has no more than 3 core features. For each feature, tell me why it made the cut and what I''d cut if I had to cut one more.
Why it works

Forces the AI into consultant mode instead of build mode. You end up with a smaller, sharper spec — and you learn what the AI thinks matters.

Getting StartedFree

Pick What to Build First

When to use: When you have too many ideas or too many half-finished projects and can't decide where to spend your time.

I''m a non-technical founder trying to decide what to build next. Here are my options: [LIST YOUR IDEAS, ONE PER LINE]. For each idea, rate it 1-10 on: (1) how much I''d actually use it, (2) how obvious the customer is, (3) how likely I am to finish it, (4) how much I''ll learn along the way. Then recommend ONE to start with and explain why. Be honest — if one is a bad idea, say so.
Why it works

Rating forces a rubric. Asking for a single recommendation forces a decision. "Be honest" gives the AI permission to disagree with you.

Design & UIFree

Make It Look Better

When to use: When the app works but looks generic or "AI-built." Use this on any page that feels bland.

Look at the current [PAGE NAME] page. It works but feels generic. I want it to look intentional and modern — like a real product, not a template. Improve the visual hierarchy, spacing, typography, and use of color. Keep the same features and content, don''t remove anything. Explain briefly what you changed and why. Don''t use generic AI-looking gradients (purple to indigo, etc).
Why it works

Constraints ("don't remove anything," "no purple gradients") stop the AI from either over-editing or defaulting to its stock aesthetic. Asking it to explain makes the changes reviewable.

Fixing & DebuggingFree

Fix This Bug

When to use: When something broke and you don't know why. Paste this before you start guessing.

Something is broken. Here''s what I expected to happen: [WHAT SHOULD HAPPEN]. Here''s what actually happened: [WHAT DID HAPPEN]. Here''s the error message if there was one: [PASTE ERROR]. Before you change any code, tell me: (1) what you think the cause is, (2) what you''d check first to confirm, (3) the smallest fix. Wait for me to confirm before making changes.
Why it works

"Wait before changing code" prevents the AI from thrashing. Asking for the smallest fix stops it from rewriting three files when one line was wrong.

Backend + LaunchFree

Add Login and Accounts

When to use: When you're ready for real users. This adds authentication cleanly on the first pass so you don't have to redo it later.

Add user accounts to my app. I need: (1) email/password sign-up and sign-in, (2) Google sign-in as an option, (3) a simple profile page where the logged-in user can see their name and email and sign out. Protect these pages from non-logged-in users: [LIST YOUR PROTECTED PAGES]. Leave the rest of the site public. Use the built-in backend, don''t ask me for API keys.
Why it works

Being explicit about which pages are protected saves a full round of rework. Naming Google sign-in up front avoids a second migration later.

Fixing & DebuggingFree

Explain This Error Like I'm 50

When to use: When you see a scary red error message and have no idea what it means.

Here''s an error I''m getting: [PASTE FULL ERROR MESSAGE]. I''m not a developer. Explain in plain English: (1) what this error actually means, (2) what likely caused it in MY app specifically (look at my code), (3) what happens if I ignore it. Don''t suggest a fix yet — I just want to understand it first.
Why it works

"Don't fix it yet" turns the AI into a teacher instead of a repair guy. You learn something. Next time you see this error you'll recognize it.

Design & UIFree

Make It Mobile-Friendly

When to use: When your app looks good on a laptop but breaks or feels wrong on a phone.

Review my entire app for mobile. Assume the user is on an iPhone with a narrow screen. Find every place where: text is too small to read, buttons are too small to tap, layout breaks, or something is hidden off-screen. List them all first. Then fix them one by one, starting with the most important. Don''t make desktop worse in the process.
Why it works

Auditing before fixing stops the AI from playing whack-a-mole. "Don't make desktop worse" is the constraint that saves you from a second round of cleanup.

Backend + LaunchFree

Write My Landing Page

When to use: When your app works but you have nothing to send visitors to.

Write and build a landing page for [YOUR APP NAME]. It''s for [YOUR TARGET USER] who struggles with [THEIR PROBLEM]. The app solves this by [YOUR SOLUTION]. Include: a headline that names the problem, a subheadline that names the outcome, one primary CTA that says [BUTTON TEXT], three benefits (not features) written as short paragraphs, one section of social proof or trust signals (leave placeholders if I don''t have any yet), and a simple footer. Write it in a warm, human voice — no marketing jargon, no "revolutionize," no "seamless."
Why it works

Benefits vs features is the difference between "we use AI" and "you'll get this done in an hour instead of a week." Banning specific words strips out the generic voice.

Getting StartedFree

What Should I Build Next

When to use: When you've shipped something and don't know what deserves your next hour of work.

Here''s what my app does today: [PASTE OR DESCRIBE]. Here''s a list of things I''ve been considering adding: [LIST 3-7 IDEAS]. For each one, tell me: (1) how much it would meaningfully change the user''s experience, (2) how much complexity it adds behind the scenes, (3) whether I should do it now, later, or never. Then recommend the single next thing worth building and why.
Why it works

A ranking makes the AI show its work. "Now / later / never" is the honest bucketing you'd never get from "what do you think?"

Already bought the Pack?

Paste your access link, or head straight to your library.

Open my library →