Free Resource
Stop prompting.
Start briefing.
The reason your AI output sounds generic isn't the tool. It's how you're asking. Our free Prompt Guardian fixes that in minutes.
Free to use · No sign-up · Works with ChatGPT
THE PROBLEM
You've been trained to Google things. AI is different.
We've spent 20+ years treating the internet like a vending machine. Keyword in, result out. AI looks the same, so we use it the same way. That's why so much AI output is useless.
You type a vague sentence, hope for the best, get something generic, give up or start fiddling. The problem isn't the model. It's the question.
Treat AI like a new colleague. Give it context, a role, constraints, and a clear task. Brief it properly and the output shifts completely.
You wouldn't ask a new hire to write a board-ready report based on a five-word shout across the office. Don't do it to your AI either.
THE DIFFERENCE IN PLAIN ENGLISH
Think of it like ordering lunch.
The same logic applies whether you're ordering a sandwich or briefing an AI model. Vague in, vague out.
THE DIFFERENCE IN PLAIN ENGLISH
Five steps to a better prompt
The Prompt Guardian does the heavy lifting. You just need a rough idea of what you want.
Click the link below. It opens directly in ChatGPT. No downloads, no sign-ups beyond your own account.
Give it your draft or a rough idea. Something like "Write a blog post about video production" is plenty to start.
The Guardian flags where the AI is likely to hallucinate, use clichés, or default to "helpful American corporate assistant" mode.
You receive a fully structured brief with Role, Context, Constraints, and Task already defined. Corporate fluff stripped. British English enforced.
Copy the engineered prompt into ChatGPT, Claude, or whichever tool you use. Watch the quality shift. That's it.
Try it now. It's free.
No theory, no courses, no faff. Just a better prompt in under five minutes.
Your data, your business
Your conversation stays private in your own ChatGPT account. KINTAL has zero access to your history or inputs.
KINTAL doesn't capture or use your inputs to train any models or tools. Nothing is stored on our end.
Use the Guardian to refine briefs for client projects. Just don't paste anything genuinely confidential into the prompt itself.
Your broader data privacy is governed by your own OpenAI and ChatGPT account settings.
Questions we’re asked all the time about prompting
-
As useful as it is, you don’t need to worry about having to learn something technical. "Prompt Engineering" is just a scary technical term for clear communication.
If you can explain a task to a junior member of staff without them coming back five times to ask what you meant, you can use AI. The problem isn't that the tech is hard; it's that we are used to Googling things with keywords.
AI needs context, constraints, and examples. The Guardian handles the structure so you can focus on the intent.
-
Because you are not asking the right questions.
Large Language Models (LLMs) are designed to predict the most likely next word. If you give them a generic instruction ("Write a blog post about leadership"), they give you the most statistically average blog post on the internet.
To fix it, you need constraints. You need to tell it who it is (Role), what it cannot say (Negative Constraints), and exactly who it is talking to - and don’t forget your own tones of voice.
If you don't, the model defaults to "helpful American corporate assistant".
-
You can never stop it 100%, but you can reduce the risk by changing how you ask.
Hallucinations happen when a model tries to fill a gap in its knowledge to please you.
To fix this:
Provide the source material: Paste the text you want it to work from.
Set a "Refusal" rule: Explicitly tell the model: "If you do not find the answer in the text provided, state 'I do not know'. Do not guess."
Our Prompt Guardian adds these safety rails automatically.
-
Because AI is probabilistic, not deterministic. It is rolling a dice for every word it chooses. This is a feature, not a bug, it’s why AI can be creative. But it is annoying when you want consistency. If you need the same output every time, your prompt needs to be rigid. You need to lock down the structure and the format. The more open the prompt, the more random the result.
-
Think of it as a pre-flight check for your ideas. It is a custom tool we built to stop us from being lazy with AI. Most people type a vague sentence into ChatGPT when we girst get started, and hope for the best.
The Guardian forces you to slow down. It reviews your draft, strips out corporate fluff (like "delve" or "leverage"), and rewrites it into a structured brief with a clear Role, Context, and Task.
And it doesn’t just write the prompt; it teaches you how to structure your thinking so you get a result you can actually use.
-
We know this is the biggest worry for businesses. For the Guardian Tool:
We (KINTAL) cannot see your chats.
We do not store your inputs, and we do not use them to train our models.
For your general AI use:
Never paste sensitive client data, PII (Personally Identifiable Information), or unreleased financial figures into a public LLM.
If you need to edit a sensitive report, replace the client name with [CLIENT] and the numbers with placeholders before you start.
Top tip, python is brilliant at redacting your documents. And you don’t need to be a software engineer to do it these days, ask AI to walk you through how to do it.
-
Yes. In fact, it is specifically for you. The "Tech Bros" already have their own complex coding workflows.
We built the Guardian for the rest of us; creatives, writers, and operators who just want to get a job done on a Tuesday morning without learning Python.