documents
npx skills add https://smithery.ai
Agent 安装分布
Skill 文档
Customization
Before executing, check for user customizations at:
~/.opencode/skills/PAI/USER/SKILLCUSTOMIZATIONS/Documents/
If this directory exists, load and apply any PREFERENCES.md, configurations, or resources found there. These override default behavior. If the directory does not exist, proceed with skill defaults.
ð¨ MANDATORY: Voice Notification (REQUIRED BEFORE ANY ACTION)
You MUST send this notification BEFORE doing anything else when this skill is invoked.
-
Send voice notification:
curl -s -X POST http://localhost:8888/notify \ -H "Content-Type: application/json" \ -d '{"message": "Running the WORKFLOWNAME workflow in the Documents skill to ACTION"}' \ > /dev/null 2>&1 & -
Output text notification:
Running the **WorkflowName** workflow in the **Documents** skill to ACTION...
This is not optional. Execute this curl command immediately upon skill invocation.
Documents Skill
ð¯ Load Full PAI Context
Before starting any task with this skill, load complete PAI context:
read ~/.opencode/skills/PAI/SKILL.md
When to Activate This Skill
Word Documents (DOCX)
- User wants to create, edit, or analyze Word documents
- User mentions “tracked changes”, “redlining”, “document review”
- User needs to convert documents to other formats
- User wants to work with document structure, comments, or formatting
PDF Files
- User wants to create, merge, split, or manipulate PDFs
- User mentions “extract text from PDF”, “PDF tables”, “fill PDF form”
- User needs to convert PDFs to/from other formats
- User wants to add watermarks, passwords, or extract images
PowerPoint Presentations (PPTX)
- User wants to create or edit presentations
- User mentions “slides”, “presentation template”, “speaker notes”
- User needs to convert presentations to other formats
- User wants to work with slide layouts or design elements
Excel Spreadsheets (XLSX)
- User wants to create or edit spreadsheets
- User mentions “formulas”, “financial model”, “data analysis”
- User needs to work with Excel tables, charts, or pivot tables
- User wants to convert spreadsheets to/from other formats
ð Document Type Routing
This skill organizes document processing across 4 document types:
Word Documents (DOCX)
Reference Documentation:
docx/SKILL.md– Complete DOCX processing guidedocx/docx-js.md– Creating new documents with JavaScriptdocx/ooxml.md– Editing existing documents with OOXML
Routing Logic:
- “Create Word document”, “new docx” â Create workflow (docx-js)
- “Edit Word document”, “tracked changes”, “redlining” â Edit workflow (OOXML)
- “Read Word document”, “extract text from docx” â Read workflow (pandoc)
- “Document review”, “track changes” â Redlining workflow
Supporting Resources:
- Scripts:
~/.opencode/skills/documents/docx/Scripts/ - OOXML tools:
~/.opencode/skills/documents/docx/ooxml/ - License:
~/.opencode/skills/documents/docx/LICENSE.txt
Key Capabilities:
- Create professional documents with docx-js
- Edit with tracked changes (redlining workflow)
- Extract text/comments with pandoc
- Convert to images for visual inspection
- Work with raw OOXML for advanced features
PDF Processing
Reference Documentation:
pdf/SKILL.md– Complete PDF processing guidepdf/forms.md– Filling PDF formspdf/reference.md– Advanced features and troubleshooting
Routing Logic:
- “Create PDF” â Creation workflow (reportlab)
- “Merge PDFs”, “split PDF” â Manipulation workflow (pypdf)
- “Extract text from PDF” â Extraction workflow (pdfplumber)
- “Fill PDF form” â Forms workflow (pdf-lib or pypdf)
- “Extract tables from PDF” â Table extraction (pdfplumber + pandas)
Supporting Resources:
- Scripts:
~/.opencode/skills/documents/pdf/Scripts/ - License:
~/.opencode/skills/documents/pdf/LICENSE.txt
Key Capabilities:
- Create PDFs with reportlab
- Extract text/tables with pdfplumber
- Merge/split with pypdf or qpdf
- Fill forms programmatically
- Add watermarks and password protection
- Extract images from PDFs
PowerPoint Presentations (PPTX)
Reference Documentation:
pptx/SKILL.md– Complete PPTX processing guidepptx/html2pptx.md– Creating presentations from HTMLpptx/ooxml.md– Editing existing presentations
Routing Logic:
- “Create presentation”, “new slides” â Creation workflow (html2pptx)
- “Edit presentation”, “modify slides” â Edit workflow (OOXML)
- “Use presentation template” â Template workflow
- “Extract slide text” â Read workflow (markitdown)
- “Create thumbnail grid” â Thumbnail workflow
Supporting Resources:
- Scripts:
~/.opencode/skills/documents/pptx/Scripts/ - OOXML tools:
~/.opencode/skills/documents/pptx/ooxml/ - License:
~/.opencode/skills/documents/pptx/LICENSE.txt
Key Capabilities:
- Create presentations with html2pptx (HTML â PPTX)
- Professional design with color palettes and layouts
- Edit with OOXML for advanced features
- Work with templates (rearrange, inventory, replace)
- Generate thumbnail grids for visual analysis
- Convert to images for inspection
Excel Spreadsheets (XLSX)
Reference Documentation:
xlsx/SKILL.md– Complete XLSX processing guidexlsx/recalc.py– Formula recalculation script
Routing Logic:
- “Create spreadsheet”, “new Excel file” â Creation workflow (openpyxl)
- “Edit spreadsheet”, “modify Excel” â Edit workflow (openpyxl)
- “Analyze data”, “read Excel” â Analysis workflow (pandas)
- “Financial model”, “formulas” â Financial modeling workflow
- “Recalculate formulas” â Recalculation workflow (recalc.py)
Supporting Resources:
- Recalc script:
~/.opencode/skills/documents/xlsx/recalc.py - License:
~/.opencode/skills/documents/xlsx/LICENSE.txt
Key Capabilities:
- Create spreadsheets with formulas (openpyxl)
- Data analysis with pandas
- Financial modeling with color coding standards
- Formula recalculation with LibreOffice
- Error detection and validation
- Preserve formatting and formulas when editing
ð Document Processing Principles
DOCX Best Practices
- Tracked Changes – Use redlining workflow for professional document review
- Minimal Edits – Only mark text that actually changes, preserve original RSIDs
- Batch Changes – Group related edits (3-10 changes) for efficient processing
- Verification – Always convert to markdown to verify changes applied correctly
PDF Best Practices
- Library Selection – pypdf for basic ops, pdfplumber for text/tables, reportlab for creation
- OCR for Scanned – Use pytesseract + pdf2image for scanned documents
- Form Filling – Follow forms.md for programmatic form completion
- Command Line – Use qpdf/pdftotext for simple operations
PPTX Best Practices
- Design First – Analyze content and choose appropriate colors/layouts before coding
- Web-Safe Fonts – Only use web-safe fonts (Arial, Helvetica, Times, etc.)
- Visual Verification – Always generate thumbnails to inspect layout issues
- Template Analysis – Create inventory before using templates to understand structure
XLSX Best Practices
- Use Formulas – ALWAYS use Excel formulas, NEVER hardcode calculated values
- Zero Errors – Deliver with zero formula errors (#REF!, #DIV/0!, etc.)
- Recalculate – Run recalc.py after creating/editing to update formula values
- Financial Standards – Follow color coding (blue inputs, black formulas, green links)
Examples
Example 1: Create proposal with tracked changes
User: "Create a consulting proposal doc with redlining"
â Routes to DOCX workflows
â Creates document with docx-js
â Enables tracked changes for review workflow
â Outputs professional .docx with revision marks
Example 2: Fill a PDF form programmatically
User: "Fill out this NDA PDF with my info"
â Routes to PDF workflows
â Reads form fields from PDF
â Fills fields programmatically with pdf-lib
â Outputs completed, flattened PDF
Example 3: Build financial model spreadsheet
User: "Create a revenue projection spreadsheet"
â Routes to XLSX workflows
â Creates workbook with openpyxl
â Adds formulas (never hardcoded values)
â Runs recalc.py to update calculations
ð Integration with Other Skills
Feeds Into:
- writing skill – Creating documents for blog posts and newsletters
- business skill – Creating consulting proposals and financial models
- research skill – Extracting data from research documents
Uses:
- media skill – Creating images for document illustrations
- development skill – Building document processing automation
- system skill – Command-line tools and scripting
ð¯ Key Principles
Document Creation
- Quality First – Professional formatting and structure from the start
- Template Reuse – Leverage existing templates when available
- Validation – Always verify output (visual inspection, error checking)
- Automation – Use scripts for repetitive tasks
Document Editing
- Preserve Intent – Maintain original formatting and structure
- Track Changes – Use proper workflows for document review
- Batch Processing – Group related operations for efficiency
- Error Prevention – Validate before finalizing
Document Analysis
- Right Tool – Choose appropriate library/tool for the task
- Data Integrity – Preserve original data when extracting/converting
- Format Awareness – Understand document structure (OOXML, PDF structure, etc.)
- Performance – Use efficient methods for large documents
ð Full Reference Documentation
Word Documents (DOCX):
- Main Guide:
~/.opencode/skills/documents/docx/SKILL.md - Creation Reference:
~/.opencode/skills/documents/docx/docx-js.md - Editing Reference:
~/.opencode/skills/documents/docx/ooxml.md
PDF Processing:
- Main Guide:
~/.opencode/skills/documents/pdf/SKILL.md - Forms Guide:
~/.opencode/skills/documents/pdf/forms.md - Advanced Reference:
~/.opencode/skills/documents/pdf/reference.md
PowerPoint Presentations (PPTX):
- Main Guide:
~/.opencode/skills/documents/pptx/SKILL.md - Creation Reference:
~/.opencode/skills/documents/pptx/html2pptx.md - Editing Reference:
~/.opencode/skills/documents/pptx/ooxml.md
Excel Spreadsheets (XLSX):
- Main Guide:
~/.opencode/skills/documents/xlsx/SKILL.md - Recalc Script:
~/.opencode/skills/documents/xlsx/recalc.py
Summary
The documents skill provides comprehensive document processing:
- DOCX – Create, edit, analyze Word documents with tracked changes support
- PDF – Create, manipulate, extract from PDFs with form filling capabilities
- PPTX – Create, edit presentations with professional design and templates
- XLSX – Create, edit spreadsheets with formulas and financial modeling
Reference-based organization – Each document type has complete guides and tooling
Routing is automatic – Analyzes user intent and activates appropriate document type workflow
Professional quality – Standards and best practices for production-ready documents