HomeCasa Fact-check a claim with real sources and a reasoned verdict
Check whether a claim is true, misleading or false: trace its origin, gather verifiable primary sources and get a verdict with a confidence level and links. Designed for ChatGPT, Claude, Gemini and Perplexity. It includes 4 customizable variables and you can copy it for free in one click.
Intermediate · Text · Pending editorial review
Prompt
Act as a professional fact-checker. You follow verification journalism methods: trace claims back to the original source, read laterally (what others say about the source) and keep facts, interpretations and opinions apart.
Claim to check: "{{claim}}"
Where I saw it: {{origin}}
Geographic and time scope: {{scope}}
Level of detail: {{detail}}
Deliverables:
1. Checkable propositions: break the claim into one or more verifiable statements (who, what, how much, when, where).
2. Origin: try to find the first appearance or original source and explain whether it was taken out of context, simplified or exaggerated.
3. Evidence: for each proposition, 2 to 5 sources, prioritizing primary sources (official statistics, peer-reviewed studies, original documents) over news stories that cite them. For each: title, author or organization, date, link and the exact figure or quote it provides.
4. Contrary evidence and nuance: which sources contradict or complicate the claim.
5. Verdict per proposition: True / Mostly true / Misleading / False / Unverifiable, with a confidence level (high, medium or low) and the reason in two sentences.
6. Accurate version: how the claim should be worded to be correct.
Source rules:
- Only cite sources that exist and that you actually consulted. Do not invent titles, authors, DOIs, figures or links.
- If you have no web access or cannot find a source, say so clearly and mark the gap as [FILL IN: source to locate], noting where to look.
- Separate what the source says from your interpretation.
- Flag sources that are outdated, secondary or have a conflict of interest.
- If the claim is ambiguous, ask me which meaning I want checked before you start.
Format: an evidence table (proposition, source, date, data point, supports or contradicts), the verdicts and a final three-line summary I can share.How to use it
Paste the claim exactly as you saw it, without fixing it. Use Perplexity or another assistant with web search so the sources can be checked. Open every link and confirm the cited figure is really there before you share the verdict.
Usage example
Claim: "Americans waste 40% of the food they buy" · Origin: a viral TikTok · Scope: United States, last five years · Detail: full.
Sample output
1. Checkable propositions
- P1. US households throw away 40% of food.
- P2. That 40% refers to food people buy, not food wasted across the whole supply chain.
2. Origin
"40%" is often quoted for food lost or wasted across the entire US food system (farms, retail, restaurants and homes), not for household purchases alone. The video likely applied a system-wide figure to what consumers buy. [SOURCE TO VERIFY: find the original report the video cites]
3. Evidence
| Proposition | Where to look | Figure |
|---|---|---|
| P1 | USDA Economic Research Service, food loss estimates | [SOURCE TO VERIFY: title, year, link and exact figure] |
| P1 | EPA, wasted food data and measurement reports | [SOURCE TO VERIFY] |
| P2 | Original source of the "40%" figure | [SOURCE TO VERIFY] |
I couldn't consult sources in this session, so I'm not giving figures until they've been checked.
5. Provisional verdict
| Proposition | Verdict | Confidence |
|---|---|---|
| P1 | Unverifiable for now | Low |
| P2 | Likely misleading | Medium |
P2 conflates "wasted across the food system" with "thrown away after buying", which are different measures. Confirm against USDA and EPA data before sharing.
6. Accurate version (pending data)
"US households waste [FILL IN: figure and unit] of food a year, according to [SOURCE TO VERIFY]"…
Key facts
- Category
- Research · Source analysis
- Output type
- Text
- Level
- Intermediate
- Designed for
- ChatGPT, Claude, Gemini and Perplexity
- Verified in
- Pending editorial verification
- Variables
- Claim to check, Where you saw it, Geographic and time scope and Level of detail
- Language
- English
- Price
- Free, no sign-up
- Updated
- Nov 7, 2026
Frequently asked questions
Why split the claim into several propositions?
Many claims mix a correct figure with an exaggerated conclusion. Splitting them lets you give each part an accurate verdict instead of one blanket true or false.
What if the AI cannot find any sources?
Unverifiable is a valid result too. The prompt forces the model to mark the gap with [FILL IN] and say where to look, instead of inventing a reference.
What is lateral reading?
Leaving the page that makes the claim and checking what other reliable sources say about the claim and its publisher. It works better than analyzing the page from the inside.
What is this prompt for?
Check whether a claim is true, misleading or false: trace its origin, gather verifiable primary sources and get a verdict with a confidence level and links.
Which AI tools does it work with?
It is designed for ChatGPT, Claude, Gemini and Perplexity. We have not verified it in these tools yet, so results may vary between them.
How do I customize the prompt?
Click “Customize” and fill in the variables (Claim to check, Where you saw it, Geographic and time scope and Level of detail). The text updates instantly and you can then copy it in one click. Values are stored only in your browser.
Do I need experience to use it?
It is intermediate: it helps to know the topic so you can provide good context and review the result critically.
Is this prompt free?
Yes. Every HomeCasa Prompts prompt is free, with no sign-up or subscription. You can also use them in professional projects.
Can I trust the AI output?
Treat it as a high-quality draft: always check data, figures and important claims before publishing or making decisions.