HOW WE
PUBLISH.
TrustSniffer publishes automated risk assessments and written research. This policy explains how that content is produced, reviewed, and held to a standard, across everything we put out, not just one product.
One standard for everything we publish.
This policy applies to every kind of content TrustSniffer makes public: website risk reports, wallet risk verdicts, the aggregate Risk Index, our research articles, and the summaries we post to our channels. Whether a result is produced automatically by our analysis engine or written up as an article, the same principles below apply.
Nothing we publish is a mystery number.
Grounded in evidence
Every verdict and material claim rests on evidence we can point to, on-chain records, page and infrastructure signals, or cited sources.
Reasons shown
A result comes with the reasons behind it, so a reader can see why, not just a score.
Attributed, not absolute
We describe what our own analysis found. We avoid blanket "scam" labels and treat assessments as probabilistic, not final proof.
Dated & revisable
Time-sensitive findings carry a date and can change as new evidence arrives. A verdict reflects what we could see at that time.
AI helps. People decide.
Articles
Research articles may be drafted with AI assistance and pass automated quality and accuracy checks, but every article is reviewed and approved by a person before it is published. Automated systems never publish on their own, and we tell you when AI assisted a piece.
Verdicts & reports
Website and wallet assessments are produced by our analysis engine from evidence, not written by an AI. AI may help phrase the plain-English explanation of a result; it does not decide the outcome. The verdict follows the evidence.
The bar for going live.
Sourcing
Claims trace to a source or to our own first-party analysis. We don't present unsupported assertions as fact.
Accuracy & limits
Figures tie to data, and we state the limits of what a result covers rather than overclaiming.
Fairness
An assessment is not an accusation of a crime. We describe risk and evidence, and we correct mistakes.
Independence
Verdicts are not for sale. Paying for a fuller report or faster access never changes a result, and we don't accept payment to alter or remove a rating.
Lines we don't cross.
No fabrication. We do not invent claims, sources, figures, or quotes. If we can't support it, we don't publish it.
No hidden sponsorship. We do not accept payment to publish, favor, or suppress a verdict or article, and any commercial relationship that affected a piece would be disclosed.
Not legal advice. Our content is informational. It is not legal, financial, or compliance advice and is not a legal finding.
No private data. We publish analysis of public websites and public on-chain activity, not private user data.
We fix mistakes in the open.
If we get something wrong, we correct it, and we don't quietly delete the record. How corrections and retractions work, and how to flag a result you believe is inaccurate, is set out in our corrections policy. The limits of what any result establishes are in our disclaimer.
See how we decide.
Read the methodology behind our verdicts, or what we support today.