How we decide what to say

Editorial policy

A security verdict is an accusation with a number attached. These are the rules that constrain what we will and will not say, and what happens when we get one wrong.

Who writes this

Veriql is built and run by one named person, Joseph, working in the open. Who that is, and why the name is on it at all, is on the about page. There is no editorial board and no anonymous staff: one person is accountable for every judgement on this site, which is the only arrangement under which “hold us to it” means anything.

What we publish

A verdict is a measurement with its arithmetic shown. Every penalty names the reading that produced it, and the full weighting is public on the methodology page. If you disagree with a score you can still use the evidence, which is the point: the number is an opinion about measurements, and the measurements are facts you can re-check on chain.

Outcomes are recorded as what was measured, never as what we think somebody did. “Pool liquidity went from $X to $Y, N hours after this verdict” is a statement anyone can verify. “The team rugged” is an accusation about people we cannot name, and we do not make it.

What we refuse to publish

Money, and where it cannot reach

Veriql earns a share of trading fees when a reader follows a trade link after a clean verdict. That creates an obvious incentive to call things clean, so the gate is absolute and mechanical: a flagged or incomplete verdict can never show a trade link, and the code refuses it rather than leaving it to judgement. The refusal to earn on scams is the only thing that makes the clean verdicts worth reading, so it is enforced in software and covered by tests.

We are not paid by any project to review it, and no project can pay to change a score, be removed from Drain Watch, or be added to it. If that ever changes it will be disclosed on this page before it happens, not after.

Sources, and what we own

Where a finding comes from a third party, the page says which checks ran. Most signals are read directly from public chain state, which is why they can be re-checked independently. Sanctions data mirrors the US Treasury Specially Designated Nationals list, a US federal government work, through a public mirror we credit openly. Licensed threat feeds are used only where the licence permits redistribution, and the ingest fails closed when a licence is missing rather than quietly shipping the data.

When we are wrong

We publish the misses. The track record counts verdicts where we warned and nothing happened alongside verdicts where we found nothing and the pool emptied, in the same cells and the same type size, and it states no success rate because a percentage between small numbers reads as evidence without being any.

A verdict we published and later disproved is deleted only when it was factually wrong at the time it was written, and the deletion is recorded. A verdict that was our honest conclusion from the data we had stays up even when a later reading disagrees, because rewriting it would be the worse dishonesty.

Corrections and disputes

If you are a project owner who believes a verdict is wrong, send the contract address and what you think we misread. We will re-run the check, and if we were wrong we will say so on the record rather than quietly adjusting a number. Contact.

What this is not

Nothing on this site is financial advice, legal advice, or a compliance opinion. A clean verdict is a statement that specific structural checks found nothing, on a specific date. Contracts get upgraded, liquidity gets pulled, and people change their minds. Do your own research and never risk money you cannot lose.