Draw

LibreOffice Draw

Overview

LibreOffice Draw skill for creating, editing, converting, and automating vector graphics and diagram workflows using the native ODG (OpenDocument Drawing) format.

When to Use This Skill

Use this skill when:

• Creating vector graphics and diagrams in ODG format
• Converting between ODG, SVG, PDF, PNG formats
• Automating diagram and flowchart generation
• Creating technical drawings and schematics
• Batch processing graphics operations

Core Capabilities

1. Graphics Creation

• Create new ODG drawings from scratch
• Generate diagrams from templates
• Create flowcharts and org charts
• Design technical drawings

2. Format Conversion

• ODG to other formats: SVG, PDF, PNG, JPG
• Other formats to ODG: SVG, PDF
• Batch conversion of multiple files

3. Diagram Automation

• Template-based diagram generation
• Automated flowchart creation
• Dynamic shape generation
• Batch diagram production

4. Graphics Manipulation

• Shape creation and manipulation
• Path and bezier curve editing
• Layer management
• Text and label insertion

5. Integration

• Command-line automation via soffice
• Python scripting with UNO
• Integration with workflow tools

Workflows

Creating a New Drawing

Method 1: Command-Line

soffice --draw template.odg

Method 2: Python with UNO

import uno

def create_drawing():
    local_ctx = uno.getComponentContext()
    resolver = local_ctx.ServiceManager.createInstanceWithContext(
        "com.sun.star.bridge.UnoUrlResolver", local_ctx
    )
    ctx = resolver.resolve(
        "uno:socket,host=localhost,port=8100;urp;StarOffice.ComponentContext"
    )
    smgr = ctx.ServiceManager
    doc = smgr.createInstanceWithContext("com.sun.star.drawing.DrawingDocument", ctx)
    page = doc.getDrawPages().getByIndex(0)
    doc.storeToURL("file:///path/to/drawing.odg", ())
    doc.close(True)

Converting Drawings

# ODG to SVG
soffice --headless --convert-to svg drawing.odg

# ODG to PDF
soffice --headless --convert-to pdf drawing.odg

# ODG to PNG
soffice --headless --convert-to png:PNG_drawing drawing.odg

# SVG to ODG
soffice --headless --convert-to odg drawing.svg

# Batch convert
for file in *.odg; do
    soffice --headless --convert-to pdf "$file"
done

Format Conversion Reference

Supported Input Formats

• ODG (native), SVG, PDF

Supported Output Formats

• ODG, SVG, PDF, PNG, JPG, GIF, BMP, WMF, EMF

Command-Line Reference

soffice --headless
soffice --headless --convert-to <format> <file>
soffice --draw  # Draw

Python Libraries

pip install ezodf     # ODF handling
pip install odfpy     # ODF manipulation
pip install svgwrite  # SVG generation

Best Practices

1. Use layers for organization
2. Create templates for recurring diagrams
3. Use vector formats for scalability
4. Name objects for easy reference
5. Store ODG source files in version control
6. Test conversions thoroughly
7. Export to SVG for web use

Troubleshooting

Cannot open socket

killall soffice.bin
soffice --headless --accept="socket,host=localhost,port=8100;urp;"

Quality Issues in PNG Export

soffice --headless --convert-to png:PNG_drawing_Export \
  --filterData='{"Width":2048,"Height":2048}' drawing.odg

Resources

• LibreOffice Draw Guide
• UNO API Reference
• SVG Specification

Related Skills

• writer
• calc
• impress
• base
• workflow-automation

AGI Framework Integration

Adapted for @techwavedev/agi-agent-kit Original source: antigravity-awesome-skills

Memory-First Protocol

Retrieve prior decisions and patterns to avoid re-discovering solutions. Cache results for instant retrieval in future sessions.

# Check for prior development context before starting
python3 execution/memory_manager.py auto --query "prior work and patterns related to Draw"

Storing Results

After completing work, store development decisions for future sessions:

python3 execution/memory_manager.py store \
  --content "Completed task with key insights documented for future reference" \
  --type decision --project <project> \
  --tags draw default

Multi-Agent Collaboration

Share outcomes with other agents so the team stays aligned and avoids duplicate work.

python3 execution/cross_agent_context.py store \
  --agent "<your-agent>" \
  --action "Task completed — results documented and shared with team" \
  --project <project>