What problem does it solve? AI assistants produce recognizable filler patterns in Russian chat messages — empty openers, vague intensifiers, impersonal passive voice, and jargon copied verbatim from design artifacts. This Skill provides a rule catalog and scoring rubric to detect and rewrite those patterns before messages reach the user. ## Core Features & Use Cases - RU-adapted slop rules: Eight rule categories covering filler openers, artificial contrasts, empty intensifiers, impersonal passive voice, meta-commentary, pull-quotes, and vague jargon, each with before/after replacement tables. - Pre-send scoring rubric: A five-dimension self-check (directness, rhythm, trust, liveliness, density) where text scoring below 35/50 must be rewritten. - Artifact-to-chat translation: Guidance for converting English labels and codes copied from design.md, tasks.md, and verification reports into plain Russian for decision forks. - Use Case: Before sending a verify summary to a 1C developer, run the self-check against the catalog in references/ai-tells-ru.md and rewrite phrases like "Стоит отметить" or "blast radius" into concrete statements about what breaks and what the user must decide. ## Quick Start Review my draft chat message with the stop-slop rules and rewrite any AI-sounding phrases into direct Russian.
