Method

The method is public. The rule engine decides the result.

AI may help read public material, but it cannot decide the verdict. Every evidence item needs a source, and scoring follows a fixed rules table.

1. Normalize the URLRemove tracking parameters and paths, keeping the main domain to reduce privacy risk.

2. Compare public sourcesTaiwan 165 public mirror data and RDAP registration data are stored separately. Sources that are not integrated are not used for automatic verdicts.

3. Show sources and appeal pathEvery report has an ID, check time, source links and a correction path.

Current coverageThe current verdict sources are mainly Taiwan official lists and public registration records, so coverage for scam sites outside Taiwan is limited. The English version can still be used, but a no-record result has lower reference value than on the Traditional Chinese version.

Cross-region risk signals

Evidence, computed score and public label are handled as separate records.

01Normalize URLKeep the main domain and remove tracking parameters.

02Compare sourcesOfficial lists, RDAP, certificates and review data stay separate.

03Apply policyThe same score can publish different labels by region.

Example Reports

These published and indexable reports show how the method, sources and correction path appear in practice.

0pz.topHigh risk