spam.known_phrase
The text matches a curated known spam or scam phrase for the applicable language.
What it looks at#
Whole-word/whole-phrase matching against a curated list of known spam and scam phrasing,
normalised the same way profanity.word is (case
folded, repeated letters collapsed) — same matcher, different list.
Which language’s list applies follows the same rule as the rest of this gate and
/v1/profanity: your declared languages, else the language /v1/spam detects with
confidence, else English. Today the curated list only covers English and Swedish; other
languages simply have no list to match against yet.
Why it matters#
Certain phrasing recurs across spam and scam messages regardless of the specific link or contact details attached to it — a weak signal on its own, but a real one.
What to do#
Low severity — one recognised phrase is suggestive, not conclusive. Combine with the
other /v1/spam checks.