Three steps to your first prompt

Briefli is an AI-powered interview tool that helps non-technical people describe what they want to build — and turns that description into a prompt so detailed and precise that AI coding tools like Claude, Cursor, or ChatGPT can actually build it on the first try.

Think of it like having a technical consultant ask you the right questions. You explain your idea in plain English, and Briefli figures out everything an AI needs to know to build it properly.

Briefli has a friendly conversation with you — no forms, no jargon, no tech knowledge required. It asks questions in plain language about what you want to build, who it's for, and what features matter most.

1. The interview is split into four short phases:
2. **Getting Started** — your big picture idea
3. **Your Project** — who it's for and what problem it solves
4. **Features & Details** — what it should actually do
5. **Almost Done** — quality, polish, and any constraints

When you're done, Briefli assembles everything into a structured, multi-phase prompt you can paste directly into your AI coding tool.

Any of them. Briefli generates a plain-text prompt, so you can paste it into:

• **Claude** (Anthropic)
• **Cursor**
• **ChatGPT** (OpenAI)
• **Lovable**
• **Bolt**
• **GitHub Copilot**
• Any other AI coding tool you prefer

The prompts are written to work well across all major tools. If you have a favourite, just paste your prompt in and go.

When most people ask an AI coding tool to "build a task management app," the result is usually generic, incomplete, or requires four rounds of back-and-forth corrections.

The reason is that AI tools work best when given very specific instructions — the kind that usually require technical knowledge to write.

Briefli handles that part. It extracts the details from you in plain English, then structures them into a prompt that tells the AI exactly what to build, in what order, with what constraints. You get the result that would normally take a technical co-founder to produce.

The free tier includes:

• **2 full interviews per month** — no credit card required
• **Full-quality prompt output** — same quality as Pro, not a watered-down version
• **Works with any AI coding tool**
• **Prompt saved** so you can come back to it

The main limit is 2 interviews per month. If you need more, or want multi-phase prompt chains and saved profile/history, that's what Pro is for.

Pro is $15/month and includes:

• **Unlimited interviews**
• **Multi-phase prompt chains** — longer, more complex builds get broken into sequential phases instead of one prompt
• **Saved profile** — Briefli remembers your tech preferences so you don't repeat yourself
• **Prompt history** — access all your previous outputs any time
• **Priority responses** — faster generation

If you're building more than 2 things a month, or your project is large enough to need multi-phase prompts, Pro pays for itself immediately in saved time.

Once your interview is complete, Briefli generates your prompt and shows it on a results page. Here's what to do next:

1. **Copy** the prompt (there's a one-click copy button)
2. **Open** your AI coding tool (Claude, Cursor, ChatGPT, etc.)
3. **Paste** the prompt and press send
4. **Review** the output — it should be very close to what you described

If the AI asks a clarifying question, answer it. If anything in the output isn't quite right, you can always go back to Briefli and refine your interview, or edit the prompt directly.

Yes. A few specifics:

• **We don't use your interview content to train AI models, and we don't sell it.**
• **We share content only with the service providers that operate Briefli** — Anthropic (the AI provider) processes your inputs to generate prompts under its standard API terms, which prohibit training on your content. Supabase stores your account and saved prompts. Stripe handles payments. Vercel hosts the site.
• **Your prompt output is saved to your account** so you can access it later, but only you can see it.
• **You can delete your account and data** at any time — see our [Privacy Policy](/privacy) for the full data-rights process.

We use anonymous analytics (PostHog) to understand how the product is used — no interview content is included in that.

Yes — the prompt is just text. Once you're on the results page, you can copy it, paste it anywhere, and edit it freely before you use it.

Most people use it as-is and get great results. But if something doesn't look right, or you want to add a detail Briefli missed, just edit the relevant section. The structure of the prompt is designed to be readable, so it's easy to find the part you want to change.

A few things to try:

1. **Re-run the interview.** Sometimes the question phrasing leads to a different angle on your idea. A second interview often produces a better result.
2. **Add more detail.** The more specific you are about features, the more precise the output. If you gave short answers, try longer ones.
3. **Edit the prompt directly.** If 90% of it is right but one section is off, just fix that section before you paste it in.
4. **Try a different AI tool.** Some tools handle certain types of prompts better than others.

If you're consistently getting unhelpful output and nothing above fixes it, email us at hello@briefli.io — we read every message.

Briefli was built by a small team who got tired of watching non-technical people struggle to communicate with AI coding tools. The gap between "I have an idea" and "I have a working product" was too large — and it didn't need to be.

We're indie builders ourselves, and we built Briefli using the same AI tools it helps you use better. We're based online and keep the team small so we can move fast and stay close to what users actually need.

None at all. That's the point.

Briefli's questions are written in plain language. You don't need to know what a "tech stack" is, what "authentication" means, or anything about databases. Just describe what you want to build the same way you'd explain it to a friend.

If a question uses a term you don't know, just say so — or skip it. Briefli will work with whatever you give it.