What can you learn about this?
You start from a domain, an IP, a website, an on-chain address — or a person's name. A name is honest, not magic: it returns SEPARATE public candidates (a Wikidata entity, a GitHub account), each kept apart and marked unverified. Same name is not the same person, and nothing is merged. No leaked database is queried, and no password or secret is ever shown.
Checking a token instead? SWOGE Scan reads what our AI has measured on tokens like it — same rule, every number with its sample size.
Findings
What crossing the facts says — a single fact is rarely a result, the crossing is. Each finding carries its pieces and their sources. Nothing here is a verdict.
Sources disagree
Two sources gave different answers on the same point. Both are shown, best-supported first — nothing was picked for you. The one that is wrong is often the finding.
Graph
Each line is a documented relation with a source behind it, not a conclusion. Hover a line to read its source. People are drawn apart from machines on purpose.
Facts
Every statement a source made, its confidence and score, and where it came from. A disputed value is flagged, not hidden. Filter with the layers above.
Sources reached
Every connector this investigation ran, how long it took, and what it returned — including the ones that timed out or refused us. This is the monitoring, in the open.
What was not done
Connectors left off for want of a key, connectors skipped in passive mode, and the limits — all part of the result.