The ChatGPT Prompt That Writes Your App Requirement Sheet

TL;DR. Paste the interview prompt below into any AI chat and answer its questions honestly for about 10 minutes. You get a requirement sheet under 400 words that any developer can quote from. Run the reality-check prompt before spending money, and the quote-comparison prompt after. If you then want a fixed number from us, the $500 Week-1 Build Audit takes the sheet as direct input, credited against the build

The prompt: paste this into ChatGPT, Claude, or Gemini

Copy the block below into a fresh chat. It works the same in ChatGPT, Claude, and Gemini. Answer its questions one at a time, honestly, and you'll have a requirement sheet a developer can actually price in about ten minutes.

You are a senior product engineer who scopes apps for a living. Interview me about my app idea, then produce a one-page requirement sheet a developer can price.

Rules for the interview:
- Ask ONE question at a time and wait for my answer. No multi-part questions.
- Ask at most 10 main questions, plus a follow-up when an answer is vague. Follow-ups do not count against the 10. Skip any question you can already answer from what I have said.
- Push back when my answer is vague. "Everyone" is not a user. "Like Uber but for X" is not a flow.
- If my scope is too big for a first version, say so and make me cut it to one core flow.

Cover, in this order: the problem in one sentence; who has it (a specific person, not a market); the ONE core flow from open to done; what they use today (a spreadsheet, WhatsApp, nothing); platform (web, mobile, both) and why; sign-in and payments (none, accounts, payments, marketplace); what it connects to; what I will honestly spend; what success looks like in 90 days.

Then output the requirement sheet with exactly these sections:
1. Problem (one sentence)
2. User (one specific person, one line) and what they use today
3. Core flow (numbered steps, open to done)
4. In scope for v1 (max 5 bullets), and out of scope until v1.1
5. Platform, sign-in, payments, integrations
6. Budget band and the 90-day success metric
7. Three open questions a developer will ask that I could not answer

Keep the whole sheet under 400 words. No filler, no feature I did not confirm, no invented numbers.

Copy this prompt

Two habits make the output dramatically better. Answer with specifics, because the model mirrors whatever clarity you give it: a vague brief in, confident padding out. And when it pushes back on your scope, let it win. Cutting v1 to one core flow is the single biggest lever on what your MVP will cost, and it's usually the right product call too.

What a good requirement sheet looks like

The seven sections cover every field of the one-page brief in how to explain your app idea to a developer, plus two things only an interview surfaces: the v1 scope cut, and the three questions you could not answer. Those three aren't a failure. They're the agenda for your first developer conversation, and walking in knowing what you don't know puts you ahead of most founders who show up with a 40-page vision deck and no answer to "who pays".

Keep the sheet under a page. A developer quoting from one page of confirmed decisions gives you a real number. A developer quoting from a vision deck gives you a discovery phase.

Why this beats "write me a PRD" prompts

Most of the prompts you'll find for this ask the model to generate a full requirements document in one shot. The output looks impressive and is worth very little, because the model doesn't know your users, your budget, or your constraints, so it invents plausible ones. You get a document full of decisions nobody made. Developers can smell these a mile off, and they quote defensively for them.

The interview format flips it: the model asks, you decide, and every line in the sheet is something you actually confirmed. The AI contributes the structure and the pushback, which is what it's genuinely good at. You contribute the facts, which only you have. That division of labor is the whole trick.

The reality-check prompt: try to kill the idea first

Before the sheet goes anywhere near a developer, run this one. It costs nothing, and it's the cheapest moment you'll ever get to find out the idea doesn't hold.

Here is my app requirement sheet: [paste your sheet]

Your job is to try to kill this idea before I spend real money on it. You get five questions. Ask them one at a time and wait for my answer, in this order: who exactly pays, why now, why me, what already solves 80% of this, and what has to be true for the 90-day metric to happen. Challenge every soft answer once.

Then give a verdict, in one paragraph: BUILD IT, TEST IT FIRST with a landing page, or SHELVE IT. Assume the idea fails until my answers prove otherwise, and say BUILD IT only on specific evidence, never on my enthusiasm. Name the single cheapest test that would change your verdict, and what result that test must show.

Copy this prompt

If the verdict is TEST IT FIRST, that's a $399 problem, not an $8,000 one: validate the idea with a landing page walks the whole play, benchmarks included. A failed test is the process working.

The quote-comparison prompt: when the estimates come back

Once you send the sheet out and quotes come back, paste them into this. It applies the same checks we use in red flags when hiring a dev agency and questions to ask an MVP development company, without you having to hold them all in your head.

Here is my app requirement sheet, followed by the quotes I received: [paste the sheet, then each quote]

Compare every quote against these checks:
- Fixed price or hourly (hourly on a defined scope is a running meter)
- Is the code in MY repository from day one
- IP assignment written into the contract
- Milestone payments gated on working demos, never 100% up front
- A named engineer I can talk to, not "our team"
- What the quote quietly excludes: app store submission, hosting setup, post-launch bug fixes

First output a comparison table: one row per check, one column per vendor, and "not stated" wherever a quote is silent on a check. Then flag every red flag by vendor and write the exact questions I should send each one before deciding. Recommend a quote only if the answer is genuinely clear from the evidence; otherwise tell me what is missing.

Copy this prompt

What to do with the sheet

Three honest routes, depending on what the sheet says. If the budget line came out under $2,000, don't hire anyone: build it yourself in a no-code tool, and no-code vs custom code maps where that road ends. If the sheet describes one tight flow and you want it built by professionals, send it to two or three shops and compare fixed quotes; hire someone to build an app has the full route map with price bands. And wherever you send it, the sheet is the artifact that stops you paying for a discovery phase you already did yourself.

Ours is one of the doors you can walk through with it: the $500 Week-1 Build Audit takes your sheet as direct input and returns a one-page build spec, a fixed price, and an honest go or no-go, with the $500 credited in full if we build. Most builds land between $8,000 and $18,000, fixed, about 8 weeks, and you own 100% of the code from day one. If the sheet says you're not ready, the audit says so too.