ru-text
Russian text quality for AI agents — neuroslop cleanup, typography, information style, editorial standards, UX writing, business correspondence. Scores text 0–1
What is ru-text?
Russian text quality for AI agents — neuroslop cleanup, typography, information style, editorial standards, UX writing, business correspondence. Scores text 0–10. 2,000+ linguistic atoms.
Overview of ru-text
What it is
ru-text is a skill for AI agents that performs Russian text proofreading. It works inside the agent and cleans text on the fly: typography, bureaucratic language, and 17 signs of machine-generated text (neuroslop). For each correction it provides a quote and the rule that triggered the change. It does not touch the user's words, style, or tone, and it will not rewrite the user's file unless explicitly asked.
Key features
- Corrects typography (quotes, dashes, non-breaking spaces) and bureaucratic phrasing.
- Detects 17 signs of machine-generated text, including empty praise, false antitheses, and filler openings.
- Scores text from 0 to 10 across five scales, with a quote and rule for each remark.
- Operates silently during agent-written text for typography only; full proofreading is triggered on demand or when enough issues accumulate.
- Respects citations, code, and foreign fragments inside a document – never rewrites them.
- Can be disabled with a single command or by removing the skill directory.
Use cases
- Proofreading emails, letters, and business correspondence.
- Cleaning up AI-generated drafts before publication.
- Improving UX writing for buttons, error messages, and empty states.
- Editing landing pages, documentation, and informational texts.
- Scoring any Russian text for quality before publishing.
Getting started
Give this phrase to your AI agent:Установи навык https://github.com/talkstream/ru-text глобально и вызывай его, когда работа идёт над качеством русского текста: вычитка, типографика, очистка от нейрослопа, редактура, UX-тексты, деловая переписка — или по прямому упоминанию ru-text.Alternatively, use the command: npx skills add talkstream/ru-text to install all three skills, or npx skillsbd add talkstream/ru-text/ru-text for just this one. See INSTALL.md for details.
License and pricing
License: MIT. Pricing: open source (free).
Based on the project's public README and repository metadata.