What can you do with it?
The/word-document-process command enables you to process Word documents (.docx) with AI while preserving all formatting, styles, tables, and structure. You can translate documents to any language, correct grammar and spelling, rewrite content, or simplify language - all while maintaining the original document’s visual appearance.
This skill uses an advanced Word Document Processor service that preserves all formatting including bold, italic, underline, fonts, colors, highlights, tables, lists, headers, footers, and page structure. Only the text content is modified by AI. The service automatically handles documents of any size efficiently.
How to use it?
Basic Command Structure
Parameters
Required:task_description- A natural language description of what you want to do. Examples:- “Translate to Spanish”
- “Fix all grammar and spelling errors”
- “Simplify this document for 8th grade reading level”
- “Make this more formal and professional”
- “Translate to French and make it more concise”
document_source- The document to process:- Upload a .docx file directly
- Reference a document from a previous step
- Provide a URL to the document
collection_id- Specify a specific collection to save the output files (both the processed document and log file). If not provided, uses your default collection.model- Choose which LLM model to use for processing:- Claude Models (default):
claude-sonnet-4-5-20250929,claude-sonnet-4-20250514,claude-3-7-sonnet-20250219 - OpenAI Models:
gpt-4.1,gpt-4o,gpt-4o-mini,gpt-4.1-mini - Gemini Models:
gemini-2.5-flash,gemini-2.5-pro,gemini-2.5-flash-lite
- Claude Models (default):
strategy- Processing strategy (auto-detected if not specified):SPARSE_CHANGES- For targeted corrections (grammar, spelling)DENSE_CHANGES- For wholesale transformations (translation, rewriting)
Response Format
The command returns:Examples
Basic Usage - Translate Document to Spanish
Advanced Usage - Grammar Correction with Model Selection
Complex Usage - Simplify with Fast Model
Rewrite for Different Tone
Specify Output Collection
What Gets Preserved
✅ All formatting is maintained:- Bold, italic, underline, strikethrough
- Font families, sizes, and colors
- Text highlighting and backgrounds
- Paragraph alignment and spacing
- Bulleted and numbered lists
- Tables with all formatting
- Headers and footers
- Page breaks and sections
- Images and charts (untouched)
Use Cases
Translation
- Translate contracts, reports, manuals
- Maintain legal document formatting
- Preserve tables and structured content
- Keep brand styling intact
Grammar Correction
- Clean up drafts before publication
- Fix spelling and grammar errors
- Maintain author’s voice and style
- Preserve document structure
Content Simplification
- Make technical docs accessible
- Create reader-friendly versions
- Simplify legal or medical documents
- Adapt content for different audiences
Tone Adjustment
- Make formal documents casual
- Professionalize informal writing
- Adapt for different markets
- Adjust for cultural context
Content Enhancement
- Add examples to existing lists
- Expand sections with more details
- Insert additional items matching existing format
- Enhance documents while preserving style
Processing Time
Document processing time depends on:- Document length: ~2-5 seconds per page
- AI model: Claude is balanced, GPT-4 is fastest, Gemini is most cost-effective
- Batch size: Larger batches = fewer API calls but longer per batch
- Task complexity: Translation is faster than rewriting
Best Practices
Choosing the Right Model
- Claude Sonnet 4.5 (default) - Best for translation and complex reasoning
- GPT-4.1 - Fastest processing, good for grammar correction
- Gemini 2.5 Flash - Most cost-effective, good for simple tasks
Batch Size Optimization
- Small documents (<5 pages): Use default (50)
- Medium documents (5-20 pages): Increase to 75-100
- Large documents (>20 pages): Consider splitting into sections
Task Selection
- Use
translatefor language conversion - Use
grammar-correctfor error fixing only - Use
simplifyto reduce reading complexity - Use
rewritefor tone/style changes with custom instructions
Advanced Features
Adding Content to Documents
When explicitly requested, you can add new content to existing documents while preserving formatting:When adding content, the skill intelligently clones the formatting from existing paragraphs to maintain consistency. Use a two-pass approach: first transform existing text, then add new content with replicated formatting.
- Adding items to existing lists
- Expanding sections with more examples
- Inserting additional paragraphs with matching style
- Complex restructuring
- Changing document layout
- Adding entirely new major sections with different formatting
Limitations
What This Cannot Do
❌ Translate text within images (requires OCR first)❌ Process embedded objects (Excel charts, etc.)
❌ Handle password-protected documents
❌ Process older .doc format (only .docx)
❌ Modify document layout or styles
Notes
- Documents must be .docx format (not .doc)
- Very large documents (>100 pages) may take several minutes
- All AI processing uses secure, private processing
- Original document is never modified
- Images and charts are preserved but not processed

