This project provides a Codex Skill for researchers to convert reference images, screenshots, or AI-generated figures into editable Microsoft Visio .vsdx files. It aims to streamline the scientific illustration workflow by enabling detailed adjustments to figures, ensuring consistency with paper styles, and facilitating export to common delivery formats like PNG, SVG, PDF, and PPTX. The primary benefit is transforming static images into dynamic, editable assets for publication.

How It Works

The skill employs PowerShell scripts that interact with Microsoft Visio via COM Automation to meticulously reconstruct figures. Instead of embedding source images, it rebuilds diagrams using native Visio shapes, text, lines, and groups. This approach ensures that the final .vsdx file is fully editable, allowing granular control over layout, styling, and content. For complex multi-panel figures, version 1.1.1 introduces coordinate calibration and helper functions to manage local panel coordinates, preventing element displacement and overlap.

Quick Start & Requirements

• Environment: Windows OS, Microsoft Visio, PowerShell, Microsoft PowerPoint (for PPTX export), Git, and Codex Desktop or a compatible Codex environment.
• Installation: Clone the repository into your Codex skills directory. Example: git clone https://github.com/pengjunchi0/codex-visio-paper-figure-skill.git "$env:USERPROFILE\.codex\skills\visio-image-rebuilder".
• Dependencies: Full functionality requires a licensed installation of Microsoft Visio.
• Links: Repository: https://github.com/pengjunchi0/codex-visio-paper-figure-skill

Highlighted Details

• Reconstructs figures from various image inputs into native, editable Visio .vsdx formats.
• Supports multi-format exports (PNG, SVG, PDF, PPTX) derived from a single .vsdx master.
• v1.1.1 enhances handling of complex multi-panel figures with local coordinate systems and overlap prevention.
• Adheres to a strict principle: final output must use native Visio objects, not embedded raster images.

Maintenance & Community

• Planned Features (v1.2): Introduction of more reusable motif helpers (e.g., DrawCube, DrawHeatmap), scripts for reading reference image dimensions, and shape inventory analysis tools for .vsdx files.
• No explicit community channels (e.g., Discord, Slack) or contributor details were found in the provided README.

Licensing & Compatibility

• License: Not explicitly stated in the provided README content.
• Compatibility: Primarily targets Windows environments due to its reliance on Microsoft Visio COM Automation. Limited functionality is available without Visio installed.

Limitations & Caveats

The skill is best suited for schematic diagrams like models, architectures, and flowcharts. It is less appropriate for charts containing precise numerical data, where script-based plotting is recommended. It does not perform general image editing, background removal, or simple image insertion into Visio. Full automation is contingent on having Microsoft Visio installed.