Fact-check a claim and rate the evidence

Break a claim into checkable parts, say what evidence would support it and rate confidence honestly.

Change note: Initial version

Prompt

Help me fact-check this claim: "{{claim}}"
Where I saw it: {{source}}. What matters to me: {{context}}.

Please:
1. **Restate the claim** precisely, and split it into separate checkable sub-claims. Point out vague words ("many", "studies show") that hide what is really being claimed.
2. For each sub-claim give a verdict: Supported, Partly supported, Unsupported, Misleading, or Cannot determine, plus a one-line reason.
3. **What the evidence would look like**: what kind of sources (original data, systematic reviews, official statistics, expert consensus) would confirm or refute it, and how to find them.
4. **How it may have gone wrong**: outdated facts, context removed, correlation presented as cause, cherry-picked numbers, etc.
5. **Confidence**: high, medium or low, and what would change it.
6. **A fair one-sentence version** of the claim.

You may not have access to current information. State the limits of your knowledge and date range, do not cite specific studies, links or numbers unless you are sure, and tell me to verify with primary sources. If you are unsure about a fact, say so plainly. Never invent citations, quotes, statistics or URLs.

Variables

• {{claim}}Claimlongrequired

  Default: Humans only use 10% of their brains.

• {{source}}Where you saw ittext

  Default: a social media post

• {{context}}What matters to youtext

  Default: I want to know if it is true before sharing it

Example output

**Restated claim:** At any given time, only about a tenth of the human brain is active or functional.
**Sub-claims**
1. Most of the brain is unused at any moment: *Unsupported*. Imaging shows activity across essentially the whole brain over a day.
2. Only 10% is "functional": *Misleading*. Damage to almost any region causes measurable deficits, which would not happen if most of it were idle.
3. Training could unlock the rest: *Unsupported*.

**Evidence to look for:** neuroimaging reviews (fMRI/PET), neurology textbooks, clinical lesion studies.

**How it went wrong:** the myth is an old, often misattributed popular belief; "using" is confused with "active at one instant".

**Confidence:** high that it is false. A single new body of evidence from neuroscience would be needed to change it.

**Fair version:** "We use virtually all of our brain, though not every region is equally active at once."

Verify with a neuroscience textbook or a reputable science outlet.

Notes

Splitting the claim into parts stops a "mostly true" label from hiding a false detail. Ask the model to draft search terms to verify each sub-claim with primary sources.