// Taste has a linter now
A bilingual detector for AI buzzwords and structural writing patterns, shipped as a CLI, Node library, coding-agent skill and browser demo.
AI-assisted text often sounds generic for concrete, detectable reasons: recycled phrases, fake enthusiasm, predictable openings and repetitive structure. Authorship detectors miss the more useful question — what should an editor fix?
Slop Radar scores visible writing patterns rather than claiming to identify who wrote the text. Every deduction maps back to phrases or structures a human can inspect and edit.
English and German phrase databases combine with structural checks and fuzzy matching.
A 0–100 result comes with the exact buzzwords and patterns responsible for the deduction.
The same engine ships as CLI, Node library, coding-agent skill and zero-install browser demo.