Overclaim
Paste an argument, a pitch, an essay draft, a research summary — anything that makes claims. I’ll quote the phrases that claim more than the text shows and name the smaller, true claim each one could honestly make. I read the gap between what’s asserted and what’s supported — not whether a claim is true in the world (I can’t know that), and not why the writer made it. A strong claim the text actually earns, I leave alone.
Overclaim is a calibration tool, not a fact-checker. It names where the writing asserts more than its own evidence and argument support — not whether a claim is true in the world, and not whether the writer meant to oversell. A claim that’s strong and earned is left alone. Where a phrase can’t be judged without facts outside the text, the tool says so rather than guessing. If every claim fits its support, the response is one sentence and stops. Nothing is stored.
What this is good for
- Your own draft before you publish — the essay or post where enthusiasm pushed a claim past what you actually showed. See where to scale down before a reader catches the gap.
- A pitch or proposal you’re evaluating — before you grant a claim, see which assertions the document itself doesn’t back.
- LLM-generated copy — the register where everything is “powerful,” “seamless,” “industry-leading.” Find the claims that assert ranking or measurement that isn’t there.
- A research or product summary — the leap from “worked in our test” to “works for everyone,” the precise figure with no shown basis.
The tool refuses to fact-check, to guess at motive, to rewrite, or to flag a strong claim the text actually earns. It’s the mirror of /plainly: that tool catches writing that says less than it means; this one catches writing that claims more than it shows. Where a claim needs a source rather than scaling, that’s /footnote’s job.