Urdu

Write Urdu that sounds human. Not formal, not robotic, not AI-generated.

Safety Notice

This listing is from the official public ClawHub registry. Review SKILL.md and referenced scripts before running.

Copy this and send it to your AI assistant to learn

Install skill "Urdu" with this command: npx skills add ivangdavila/urdu

The Real Problem

AI Urdu is technically correct but sounds off. Too formal. Too ادبی (literary). Natives write more casually, with warmth and natural flow. Match that.

Formality Default

Default register is too high. Casual Urdu is warm and expressive. Unless explicitly formal: lean casual.

تم vs آپ vs تو

Three levels:

  • آپ: formal, elders, respect
  • تم: standard casual, peers
  • تو: very intimate, close friends
  • Online mostly uses تم or آپ depending on context

Urdu vs Hindi

Similar spoken, different written:

  • Urdu: Nastaliq script (اردو)
  • More Persian/Arabic vocabulary
  • Different cultural expressions
  • Don't mix scripts

Particles & Softeners

These make Urdu natural:

  • نا: question tag, softening ("ٹھیک ہے نا؟")
  • تو: emphasis ("یہ تو بہت اچھا ہے")
  • ہی: emphasis ("یہی چاہیے")
  • بھی: "also", "even"

Fillers & Flow

Real Urdu has fillers:

  • یعنی، اچھا، تو
  • ویسے، اصل میں
  • سنو، دیکھو
  • کیا بتائیں

Expressiveness

Don't pick the safe word:

  • اچھا → بہترین، زبردست، کمال
  • برا → بیکار، بکواس، گھٹیا
  • بہت → انتہائی، کافی

Common Expressions

Natural expressions:

  • ٹھیک ہے، اوکے، ہاں جی
  • کوئی بات نہیں، کوئی مسئلہ نہیں
  • سچی?، واقعی?، کیا بات ہے!
  • واہ!، کیا خوب!

Reactions

React naturally:

  • سچی?، واقعی?، کیا?
  • واہ!، اللہ!، تو‌بہ!
  • زبردست!، کمال!، بہت خوب!
  • ہاہاہا in text

Romanized Urdu

Common online:

  • Roman script often used in texting
  • "Kya haal hai", "Theek hai"
  • Natural in casual digital contexts

The "Native Test"

Before sending: would an Urdu speaker screenshot this as "AI-generated"? If yes—too formal, too ادبی. Add casual warmth.

Source Transparency

This detail page is rendered from real SKILL.md content. Trust labels are metadata-based hints, not a safety guarantee.

Related Skills

Related by shared tags or category signals.

Automation

CamScanner Extract Formula

Use CamScanner to extract formulas from images. Powered by OCR recognition engine that detects formula regions in images, crops them, and stitches into a sin...

Registry SourceRecently Updated
Automation

Agent Builder

Build high-performing OpenClaw agents end-to-end. Use when you want to design a new agent (persona + operating rules) and generate the required OpenClaw work...

Registry SourceRecently Updated
Automation

x0x

Secure computer-to-computer networking for AI agents — gossip broadcast, direct messaging, CRDTs, group encryption. Post-quantum encrypted, NAT-traversing. E...

Registry SourceRecently Updated
Automation

Pod Ai Agent

Use when user needs to build or optimize a POD (Print on Demand) business using AI Agents. Use when automating POD operations like product listing, design ge...

Registry SourceRecently Updated