ru-text

ru-text is an open-source plugin that brings Russian text quality rules directly into AI coding agents. It provides ~1,040 independently formulated rules across 7 domains:

— Typography: guillemets, em dashes, non-breaking spaces, digit grouping — Information style: 97 stop-words, reader-first structure, facts over adjectives — Editorial standards: 57 comma traps, pleonasms, capitalization — UX writing: 51 button labels, error messages, empty states, forms — Business correspondence: email structure, messenger etiquette, tone

The plugin auto-activates when the agent writes or edits Russian text — no manual triggering needed. Rules are organized into 9 reference files that load on demand, keeping the context window lean.

Works with Claude Code, Codex CLI, Gemini CLI, Cursor, OpenClaw, and Notion. One codebase, seven platforms.

All rules are independently formulated based on 16 canonical Russian-language sources (Ilyakhov, Gorbunov, Milchin, Nora Gal, and others). No verbatim quotes, full attribution.

Includes /ru-check for manual text audits and /ru-score for a 0.0–10.0 quality score across 5 dimensions.

Free, MIT licensed. No telemetry, no data collection.