Method
How we check
Every check follows the same route, and every card shows its work: the claims, the verdict, the reasoning, and the sources. This page explains each step — and what a verdict does and does not mean.
Reading the material
First the material is read in full: video is transcribed, on-screen text is captured, and factual claims are separated from opinion, satire, and framing. Understanding what a video actually says is half the work — viral posts often misquote their own footage.
Checking the claims
A quick check tests the claims against the model's knowledge and flags what would need research. A deep check adds live web research and looks for sources in three places: the language of the material, English, and the language of the country where the events happened — the same story often reads differently across those three. When credible sources genuinely disagree, the verdict says so instead of picking a side.
The verdict scale
The central claims hold up against the sources found.
Part holds up and part does not, or credible sources genuinely disagree. The card carries both versions.
The central claims contradict the sources found, and the card says which ones.
Opinion, satire, personal experience. The bot says so instead of inventing a verdict.
Dates and freshness
Facts age. Every card and every shared page carries the date of its check. A recent verdict for the same content can be reused — the card says so when it is — but an old one, or one about a fast-moving story, is checked again rather than repeated.
Made by models, read by you
Checks are performed by automated models, not by human editors — that is why an answer arrives in minutes. It also means a check can be wrong: a model can misread a claim, and sources themselves can be outdated or contradict each other. That is why every card shows its sources. Follow them before acting on a verdict: a verdict is an assessment with receipts, not a pronouncement.
Corrections
Think a verdict is wrong? Send the check — or its public link — to @ivn_m on Telegram or @zwz5v on Instagram, or to motya186@gmail.com. It is re-reviewed against current sources, and you get an answer with the outcome. A verdict that does not survive review is corrected: being correctable is part of being a fact-checker.