MCP server for nutq: Tajik, Persian, Dari and Uzbek script conversion and text normalization for AI assistants
Wellknown found it in public sources; nobody has proven control of it yet. Claiming takes one click if the repository is under your GitHub account, or a small file on your domain otherwise. Verified owners get the badge, 15-minute checks, status alerts, edits that outrank crawled data, and a ranking boost.
Agents can do it too: POST https://wellknown.network/api/v1/claims with {"agent":"nutq-mcp","method":"well_known_file"} — machine-readable steps at claim.json, guide at /docs/claim.
Everything here was measured by our prober or read from a registry. Nothing is self-reported.
Attributed to the source that supplied each field. Treated as claims, not facts.
# nutq-mcp An [MCP](https://modelcontextprotocol.io) server that gives AI assistants (Claude, and any other MCP client) reliable tools for **Tajik, Persian, Dari and Uzbek** text, built on [nutq](https://github.com/BAHROMTJ/nutq). Language models often misspell Tajik in Perso-Arabic script, mix up Uzbek oʻ/gʻ, or get number words wrong. With this server, the assistant calls tested code instead of guessing. ## Tools | Tool | What it does | Example | |---|---|---| | `convert_script` | Changes the writing system | `Забони тоҷикӣ` → `زبان تاجیکی`; `O'zbekiston` → `Ўзбекистон` | | `normalize_text` | Spells out numbers, dates, times, money, %, phone numbers, units and abbreviations for TTS | `Нархи 25 $ аст` → `Нархи бисту панҷ доллар аст` | | `number_to_words` | Writes a single number in words (cardinal or ordinal, any script) | `4, uz, ordinal, Cyrl` → `тўртинчи` | | `clean_text` | Fixes characters only (look-alike letters, ي/ك, apostrophes, ZWNJ) | `o'zbek` → `oʻzbek` | | `list_languages` | Lists languages, scripts and status | `tg`, `fa`, `fa-AF`, `uz` | All tools are read-only and work offline. ## Install Requires Python 3.10+. With [uv](https://docs.astral.sh/uv/), there is nothing to install by hand. **Claude Code** ```bash claude mcp add nutq -- uvx nutq-mcp ``` **Claude Desktop** — add this to `claude_desktop_config.json`: ```json { "mcpServers": { "nutq": { "command": "uvx", "args": ["nutq-mcp"] } } } ``` Without uv: `pip install nutq-mcp`, then use `"command": "nutq-mcp"`. Then ask, for example: *"Write «Забони тоҷикӣ ширинтар аз шакар» in Persian script"* or *"Prepare this Uzbek text for speech synthesis: 2026-yil 12-mart, 15% chegirma"*. ## Development ```bash pip install -e ../ -e . python -m unittest discover -s tests ``` The tests start the server in-process through a real MCP client, and also as a subprocess over stdio. MIT license.
Mapped onto the structured taxonomy from declared text and observed tool names. Confidence shown for derived entries.
Every source is kept verbatim. Field changes are logged as events.