translate-urdu

name: translate-urdu description: Translates Humanoid AI Studio curriculum content to Urdu (RTL) or other supported languages. Code blocks, ROS 2 package names, topic names, and CLI commands are always preserved in English. license: MIT allowed-tools: gemini_generate metadata: author: Ayesha Khalid version: "1.0.0" category: education domain: multilingual primary_target_language: ur rtl_languages: "ur,ar,fa,he"

Translate Urdu Skill

Purpose

Make the Humanoid AI Studio curriculum accessible to Urdu-speaking learners by translating prose content while keeping all technical identifiers intact and renderable in RTL layout.

Instructions

1. Parse the input — split content into:
   • Translatable blocks: regular prose paragraphs, headings, callout text, list items
   • Preserved blocks: fenced code blocks (```), inline code (`), ROS 2 package names, topic names (/cmd_vel, /odom), file paths, CLI commands
2. Build the translation prompt:

   Translate the following educational content to {target_language}.
   Rules:
   - Translate all prose naturally and fluently
   - Do NOT translate anything inside triple backticks (```) or single backticks
   - Do NOT translate ROS 2 package names, topic names, or file extensions
   - Do NOT translate: True, False, None, import, def, class, return
   - For Urdu: output must be right-to-left (RTL). Add metadata flag: dir=rtl
   
   [CONTENT]
   {content}
   

3. Return structured response:

   {
     "translated_content": "...",
     "target_language": "ur",
     "rtl": true,
     "preserved_blocks": ["code_block_1", "code_block_2"]
   }
   

Supported Languages

Composition with Personalization

When a learner requests both personalization AND translation simultaneously:

1. Run personalize-chapter first (adapt to learner profile)
2. Run translate-urdu on the personalized output
3. Preserve all code blocks through both steps

Frontend Integration Note

For Urdu and other RTL outputs, the Docusaurus frontend must apply dir="rtl" and lang="ur" to the content container. The backend response includes "rtl": true in metadata to signal this requirement.

Error Handling