The model returns a structured assessment, not a verdict. Use it to challenge your own read of the source. Do not use it to grade people, posts, or organizations as good or bad.
REMINDER: Output is an analytical aid, not finished intelligence. The score reflects what the model can see in the input you paste. Short, decontextualized, or excerpted sources receive lower confidence by design. Do not paste classified material or PII you cannot lawfully share.