ForkFree

Prompts / Writing

Stress-Test Your Argument With a Sceptic

You get a ranked list of the weakest points in your argument, each paired with the best opposing view and a practical fix. It works because the AI is given a specific adversarial role and told to be fair, so it critiques instead of agreeing with you.

Worked in our testThe reply followed nearly every step and the format, with a one-sentence restatement, three ranked weaknesses, strong counterviews, concrete fixes, a closing line and no invented facts. It broke the length limit once, because the "How to fix it" section under Weakness 1 is a single sentence and not two or three. It also never explicitly labels which points are matters of opinion, apart from a passing "present it as your judgement".

The prompt

Act as a sharp but fair-minded sceptical reader who has no stake in my success and wants the truth.

Task: Challenge the argument below so I can strengthen it before I publish it to {target_audience}.

Argument:
{my_argument}

Steps:
1. Restate my main claim in one sentence so I know you understood it.
2. Find the three most serious weaknesses, such as gaps in evidence, hidden assumptions or ignored counterpoints. Rank them from most to least damaging.
3. For each weakness, explain why a reasonable person might disagree with me, using the strongest version of their view.
4. Suggest one concrete change to my wording or evidence that would answer each weakness.
5. Do not flatter me or invent facts. If something is a matter of opinion, say so.

Output: A one-sentence restatement, then three numbered sections headed "Weakness", "Why someone would disagree" and "How to fix it". Keep each section to two or three sentences. Finish with one line stating which fix to make first.

Change the parts in curly braces. Use it on a near-final draft, such as a blog post, pitch or proposal, before anyone else reads it.

What happened when we ran it

We sent

Act as a sharp but fair-minded sceptical reader who has no stake in my success and wants the truth.

Task: Challenge the argument below so I can strengthen it before I publish it to small business owners who read our monthly newsletter and are wary of new software.

Argument:
Every small business should move its bookkeeping to cloud accounting software this year. Spreadsheets cause errors and waste hours each month. Cloud tools update automatically and let your accountant see the figures in real time. Businesses that stay on spreadsheets will fall behind competitors who make faster decisions.

Steps:
1. Restate my main claim in one sentence so I know you understood it.
2. Find the three most serious weaknesses, such as gaps in evidence, hidden assumptions or ignored counterpoints. Rank them from most to least damaging.
3. For each weakness, explain why a reasonable person might disagree with me, using the strongest version of their view.
4. Suggest one concrete change to my wording or evidence that would answer each weakness.
5. Do not flatter me or invent facts. If something is a matter of opinion, say so.

Output: A one-sentence restatement, then three numbered sections headed "Weakness", "Why someone would disagree" and "How to fix it". Keep each section to two or three sentences. Finish with one line stating which fix to make first.

The model replied

**Restatement:** You claim that all small businesses should switch from spreadsheets to cloud accounting software this year, because it reduces errors, saves time, and keeps them competitive. --- **1. Weakness:** The claim is universal ("every small business") but the support is generic. A sole trader with twenty transactions a month, or a business with unusual needs, may gain little, and you offer no evidence on who benefits most. This is the most damaging gap because one plausible counterexample undermines the whole claim. **Why someone would disagree:** A well-built spreadsheet that is reviewed regularly can be accurate, cheap, and fully understood by its owner. For a very simple business, subscription fees, setup time, and a learning curve may outweigh the benefits, and "this year" is arbitrary pressure. **How to fix it:** Replace "every small business" with a conditional, such as "most small businesses that issue invoices, handle VAT or sales tax, or share books with an accountant," and say who can reasonably wait. --- **2. Weakness:** The competitive claim ("will fall behind competitors who make faster decisions") is an unsupported causal leap. It assumes bookkeeping speed is what limits decisions, and that cloud software produces better decisions rather than just faster data. This reads as fear-based persuasion, which wary readers will notice. **Why someone would disagree:** Most small businesses win or lose on product, pricing, service, and cash, not on how quickly the ledger updates. Fast figures are useless if the owner doesn't read them, and plenty of businesses thrive on monthly spreadsheets. **How to fix it:** Cut the competitor line or replace it with a specific, checkable benefit, such as "seeing your cash position and unpaid invoices today rather than at month-end." If you cite a study or survey, name the source; otherwise present it as your judgement. --- **3. Weakness:** You ignore the costs and risks that worry your readers: subscription fees, migration effort, data security, dependence on a vendor and internet access, price rises, and the fact that cloud tools also produce errors when set up or used badly. Listing only benefits makes the piece look like a sales pitch, and "update automatically" is itself a double-edged claim, since interface changes can irritate users. **Why someone would disagree:** For a cautious owner, the risk of mis-importing historical data, being locked into a vendor, or paying more each year is concrete, while the benefits are abstract. They may also reasonably note that spreadsheet errors come from lack of controls, which software does not automatically supply. **How to fix it:** Add a short, honest paragraph on costs and risks, with a practical mitigation for each (for example, run both systems in parallel for one month, export your data regularly, and ask your accountant which tool they support). Acknowledging the downsides will make your recommendation more credible. --- **Fix first:** Weaken the universal claim in Weakness 1, since it sets the credibility of everything that follows.

Run on Claude Sonnet 5.5, 02 Nov 2026. Automatic check: 4 out of 5.

Mentioned for

ChatGPT, Claude, Gemini

Where it came from

Found on Web, shared by Yash Patel (XDA Developers), 31 Oct 2026: 5 unusual AI tricks I use to get better results than a single prompt ever could.

This is our own version of the idea, written and tested by us. The original may differ.

More writing prompts

Writing Worked in our test
Turn a One-Off Prompt Into Template

You get a clean, reusable template with a small set of well-named placeholders and example values, ready to save and reuse. It works because the AI is made to separate what stays fixed from what changes before it rewrites anything.

Found on GitHub, shared by yarin-zhang
Writing Partly worked
Turn Dictated Rambling Into Finished Work

It converts a messy spoken ramble into a tidy brief and a finished draft in one go. Separating clean-up, assumptions and writing stops the AI inventing details and lets you see what it understood.

Found on Web, shared by Mahnoor Faisal (XDA Developers), 11 Oct 2026
Writing Worked in our test
Handle Vague Requests With Assumptions

You get a useful first draft straight away, plus the questions that would make it better and a list of the assumptions behind it. It works because the AI is made to expose its guesses instead of hiding them, and to still deliver something usable.

Found on Web, shared by Mahnoor Faisal (XDA Developers), 29 Oct 2026