Tomstore's token-based prompt generator: how it works

If you've already used ChatGPT, Claude or Gemini for work, you probably know the problem: the first prompt you write is rarely the right one. You need a role, context, examples, constraints — and writing it from scratch every time takes time you often don't have. Tomstore's prompt generator, available at tomstore.it/genera, exists for exactly this: turning a plain-language request into a professional prompt ready to paste into your favourite AI assistant.

What the prompt generator is

It's an online tool that builds a prompt for you, complete with a role assigned to the AI, fields to fill in with your specific details, and an anti-hallucination rule that reduces the risk of the AI inventing numbers, regulations or unverifiable facts. You don't need to know prompt engineering techniques: describe what you need in your own words, and the generator does the rest.

How it works, step by step

1. Choose your area

When you open the tool you select the area that applies to you: Legal, Tax/Accounting, Technical/Surveying, Real Estate, Business/Communication, or Other if your case doesn't fit the previous categories.

2. Describe what you want to achieve

Write a few lines describing your goal: for example "a formal notice for a client who hasn't paid" or "a monthly report for the accountant". No particular formality is required.

3. Generate, copy and use

The generator returns a complete prompt with the fields to fill in clearly highlighted. Copy it and paste it into your trusted AI assistant: the result will be more relevant and easier to check than a request written on the fly.

Tokens: how payment works

• 1 token equals 1 prompt generation.
• You pay once, with no subscription and no automatic renewal.
• The link to recover your remaining tokens also arrives by email, so you can use it on any device.
• Since this is a digital product delivered immediately, once a generation has started the right of withdrawal no longer applies to that specific generation, as clearly stated before purchase.

Why a structured prompt makes the difference

A generic prompt almost always produces a generic answer. A prompt with a defined role (for example "act as an accountant drafting a communication for a client"), fields filled in with the real details of the case, and an explicit rule forbidding the AI from inventing missing information, produces output much closer to what you actually need for work. If you want to see more examples of this approach applied to a specific field, you'll find ideas in our article on how to structure professional AI prompts. If you'd rather start from ready-made collections for your field, take a look at the Tomstore shop.

Frequently asked questions

Do I need prompt engineering experience to use it?

No. The generator is designed for people who don't have the time or inclination to learn advanced techniques: just describe your goal in plain language.

What happens if I don't use all the tokens I bought?

They remain available on your account: you can retrieve them at any time via the link sent by email, even from a different device than the one you purchased with.

Does the generated prompt work with any AI assistant?

Yes, the prompt is plain text designed to be pasted into ChatGPT, Claude, Gemini or any other conversational assistant you normally use.

What does "anti-hallucination rule" mean?

It's an instruction automatically included in the prompt that explicitly asks the AI not to invent data, regulatory references or numbers when it doesn't know them for certain, reducing the risk of plausible-sounding but wrong answers.