Student guide

What it gives you

This toolkit helps your AI act more like an academic tutor.

It provides focused educational support for improving your work.

It is designed to stop AI “helpfully” replacing your ideas, your voice and you in your work.

It is designed to stop AI producing endless scrolling lists of bullet points.

It helps you avoid the risky shortcut where AI does too much and you end up with work you cannot honestly claim as your own.

It helps nurture your education by making your writing better and helping you think better. It can help you see what is unclear, understand mistakes, test your argument and plan your next revision. It should not make the main decisions for you or produce finished work to submit.

Getting started

The basic approach is that you have to try something first. Draft a sentence, paragraph, section or plan yourself.

Then choose a library for your task: either a focused mini library or the master library. Download it from this site and upload the file to your AI. The mini libraries are the Writing Tutor (WT), Structure Tutor (ST), Academic Thinking Tutor (AT), Research Proposal Tutor (RP) and Study Workflow Tutor (SW). More details of what is in each are on the Tools page.

The AI should show you a menu. After you choose an option, the AI will ask you to type, paste or upload something you have already been working on.

After you get feedback, revise and edit your work to make your own changes.

If your course asks for an AI-use statement, you can use the SW3 tool to produce a transparent record of what you did. Make sure you know the AI policy for your course.

If the AI summarises the toolkit file instead of showing the menu, type: Do not summarise the file. Use it as instructions and show the menu.

Use it again and again across one piece of work

The toolkit is not just for one-off fixes. It is designed to be used repeatedly as a piece of work develops, from first plan to final revision.

You work in small chunks at any one moment — a sentence, a paragraph, one mistake type — because focused feedback is more useful than feedback on everything at once. But those small moves add up across the life of an essay, report or dissertation.

Early on, you might use a planning or thinking tool to test whether an idea, question or brief is workable. While drafting, you might check one section at a time. After feedback, you might turn comments into a staged revision plan. Near the end, you might tighten writing you have already produced yourself.

Small focused turns are the method. Developing the whole piece over time is the point.

One thing to know: the AI may not remember your earlier sessions. If a mistake keeps recurring, it helps to keep your own short note of the patterns you are working on, and paste relevant earlier feedback back in when you return.

The AI will not be perfect — stay in charge

AI tools are not predictable software. The same prompt can give different answers, and the AI can be confident and wrong, drift away from the tool you chose, or quietly start rewriting your work for you. This is true of every AI tool, not just this one. The toolkit reduces these problems but cannot remove them.

So treat yourself as the person in charge, not the AI. If it starts rewriting your work, type prompt to return to the menu, or tell it to use only the named tool. If feedback feels too strong, too vague, or wrong, say so and ask it to explain. You are allowed to disagree.

Always check anything factual, especially references and sources, against your own course guidance. Do not assume a citation the AI produces is correct.

Questioning the feedback is part of the learning, not a sign the tool has failed.

Push back

AI feedback is not always right, so push back. If the feedback does not fit what you meant, or you think it is wrong, say so. Ask it to reread your point. Consider checking with a human tutor if you are unsure.

If you are stuck

If you get stuck at any stage, say “I’m stuck.” The tutor should step back and help you find a manageable next move. If it can see why you are stuck, it should say so and offer help with that. If not, it should ask a short question.

What the feedback should be

Good feedback should be readable and manageable. It should help you understand one or two important problems, not bury you in a long list of things to fix.

Good feedback should protect your meaning. The toolkit should not make your work sound more academic by changing your ideas. If a suggested word may change the meaning, it should explain the difference and ask you to choose.

Good feedback should lead to your next move. The best response is not a finished answer. It should help you decide what to try next in your own writing.

Good feedback can be questioned. If the AI misreads your paragraph, tell it. A useful tutor response should take your correction seriously, not simply defend its first answer.

Privacy and responsibility

Use the toolkit to learn from your own work, and follow your course rules on AI use. You remain responsible for the final wording, claims, evidence and submission.

Be careful with anything private or about other people. Do not paste or upload names, student numbers, email addresses, interview transcripts, placement notes, client details, case studies, unpublished research, marks, feedback about another student, or confidential material unless you have checked your course, research ethics or institution rules.