Visio Image Rebuilder logo

Visio Image Rebuilder

CommunityPopular
pengjunchi0
visio-image-rebuilder

Rebuild or restyle editable Microsoft Visio diagrams from reference images and existing .vsdx files, then export deliverables as .vsdx, PNG, SVG, PDF, or PPTX. Use when the user asks Codex to open Visio, recreate a diagram from a PNG/JPG/screenshot/reference image, match a scientific model figure, update colors/typography/layout in a .vsdx, calibrate panel coordinates to avoid shifted or overlapping submodules, or make publication/presentation outputs while preserving Visio native editable shapes rather than embedding the reference image as a flat picture.

Overview

Publisherpengjunchi0
Repositorycodex-visio-paper-figure-skill
Skill namevisio-image-rebuilder
Stars
1K
Forks
48
Bundled files
5
LicenseMIT
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • 5 bundled files

    Scripts, templates, and references the model can read while it works. Files are read-only and never executed.

  • Open source

    Published by pengjunchi0 on GitHub. Read the source before you install it.

Installation

Install the Visio Image Rebuilder AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
    https://github.com/pengjunchi0/codex-visio-paper-figure-skill/tree/main
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/pengjunchi0/codex-visio-paper-figure-skill.git \
  .claude/skills/visio-image-rebuilder
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Visio Image Rebuilder in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Visio Image Rebuilder on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Visio Image Rebuilder is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

Visio Image Rebuilder

Core Rule

Recreate the reference as editable Visio content. Do not satisfy a rebuild request by inserting the whole reference image into the page. Embedding the reference image is only allowed as a temporary locked tracing layer if it is removed or hidden before delivery and the final .vsdx remains native shapes, text, connectors, and groups.

Treat .vsdx as the source of truth. Export PNG, SVG, PDF, or PPTX only after the Visio page has been rebuilt or restyled and checked for native editability.

Workflow

  1. Inspect inputs.

    • Confirm paths for the target .vsdx, reference image, requested output formats, and output directory.
    • Export the current Visio page to PNG before editing when a target file already exists.
    • Inspect the .vsdx package for pages, media entries, and shape counts.
    • Back up the target file before any write.
    • If Visio is already open, close only the target document or ask before terminating a stuck process.
  2. Decode the reference image.

    • Identify page orientation, panel boundaries, module colors, captions, text hierarchy, arrows, dashed lines, and repeated motifs.
    • Build an object inventory: containers, titles, process boxes, icons, charts, graphs, equations, connectors, captions.
    • Calibrate the canvas first, then calibrate each major panel or subpanel with explicit top-left bounds or four corner points.
    • Draw panel internals in panel-local normalized coordinates whenever the figure has dense multi-panel content.
    • Decide whether the task is a full rebuild, color/style transfer, local edit, or export-only job.
    • For dense scientific figures, first create a coarse panel map, then draw panel internals. Do not start with small decorative details.
  3. Prefer Visio automation for native edits.

    • Use COM automation on Windows when Visio is installed.
    • Use DrawRectangle, DrawOval, DrawLine, Page.Import only for small source assets, and shape cells such as FillForegnd, LineColor, LineWeight, Char.Size, Char.Color, Rounding.
    • Use explicit coordinates and IDs for fragile edits.
    • Keep grouped structure meaningful: major panels, submodules, repeated blocks, legends.
    • For nested modules, use helpers such as RectRel, TextRel, LineRel, and OvalRel so child shapes are constrained by their calibrated parent panel.
  4. Use package XML edits only for narrow, deterministic changes.

    • XML patching is appropriate for recoloring existing shapes, replacing font tables, or changing known cell values.
    • Preserve Visio XML ordering: shape-level Cell nodes should be before Section, Text, or child Shapes.
    • Avoid rebuilding complex geometry by raw XML unless COM automation is unavailable.
  5. Export requested formats from the verified Visio source.

    • Use scripts/visio_page_tools.ps1 for export-only jobs.
    • Use scripts/visio_rebuild_scaffold.ps1 with -ExportFormats for rebuilds that should immediately create deliverables.
    • Prefer SVG for vector web/manuscript handoff, PDF for review/print, and PPTX for presentation decks.
    • For PPTX, use PowerPoint COM when available; the generated slide contains the Visio page render, usually inserted from SVG.
  6. Verify without overtrusting a single signal.

    • Export at least one preview after editing when possible.
    • Inspect the .vsdx package to confirm that no full-size reference PNG/JPG was left in visio/media.
    • Check shape count and representative text labels.
    • Check that major panels do not overlap and that child shapes stay within their intended panel bounds.
    • Check every requested output file exists and is non-empty.
    • If Visio automation hangs, stop safely, close the document if possible, and report whether the file was actually modified.

Implementation Pattern

For full rebuilds, generate a script that:

  • Opens the target .vsdx with Visio COM.
  • Saves a timestamped or descriptive backup.
  • Clears or duplicates the page depending on user preference.
  • Sets page size to match the reference aspect ratio.
  • Draws native shapes in top-left reference coordinates converted to Visio coordinates.
  • Defines calibrated panel bounds for complex regions and uses panel-local coordinates for their internals.
  • Adds reusable helpers for rectangles, text boxes, ovals, lines, arrows, mini charts, graph nodes, and image-like stacks.
  • Saves the document and exports requested formats.

Start from scripts/visio_rebuild_scaffold.ps1 when building a full reconstruction script. Copy it into the workspace and customize the Draw-ReferenceFigure function rather than editing the skill copy.

For style transfer, generate a script that:

  • Reads existing shape IDs, text, approximate geometry, fill, and line colors.
  • Maps known modules to target palettes by text and group context.
  • Applies fills, borders, text colors, line patterns, and font changes to existing shapes.
  • Avoids repositioning unless the user asks for layout changes.
  • Exports only after the .vsdx has been saved and inspected.

For export-only requests, run:

powershell
powershell -ExecutionPolicy Bypass -File scripts\visio_page_tools.ps1 `
  -VsdxPath "C:\path\figure.vsdx" `
  -ExportFormats svg,pdf,pptx `
  -OutputDir "C:\path\exports" `
  -InspectPackage

For rebuild plus multi-format export, run:

powershell
powershell -ExecutionPolicy Bypass -File scripts\visio_rebuild_scaffold.ps1 `
  -VsdxPath "C:\path\figure.vsdx" `
  -PageW 16 `
  -PageH 9 `
  -RefW 1600 `
  -RefH 900 `
  -PreviewPath "C:\path\exports\figure.png" `
  -ExportFormats svg,pdf,pptx `
  -OutputDir "C:\path\exports"

Safety Checklist

  • Back up before writing outside the workspace.
  • Close any open Visio document that locks the target file before direct package edits.
  • Never delete or revert unrelated user files.
  • If a previous attempt embedded the reference image, restore from backup or remove the image shape before continuing.
  • Tell the user clearly whether the final file is native editable shapes or a flat embedded image.
  • Tell the user when PPTX export is a rendered slide rather than native PowerPoint shapes.

Acceptance Criteria

A Visio rebuild is acceptable only when:

  • Main panel positions, flow direction, captions, and module hierarchy match the reference at first glance.
  • Major panels are aligned to calibrated bounds, with no obvious submodule drift, collision, or cross-panel overlap.
  • Text remains editable and uses a consistent academic font, usually Times New Roman.
  • Repeated motifs are represented with reusable native shapes rather than pasted raster crops.
  • The final .vsdx package has no full-page raster reference image in visio/media.
  • Requested PNG/SVG/PDF/PPTX deliverables were exported from the saved .vsdx and are non-empty.
  • A preview export or package inspection was performed, or the final response explicitly states why verification was skipped.

Useful Resource

Use scripts/visio_page_tools.ps1 for common inspection, backup, preview export, multi-format export, and package checks. Use scripts/visio_export_formats.ps1 when a custom task script needs reusable export helpers. Use scripts/visio_rebuild_scaffold.ps1 as the starting point for native-shape drawing scripts; it includes both global top-left helpers and panel-local calibrated helpers. Read references/rebuild-guidelines.md when a task requires a full figure reconstruction or careful one-to-one scientific diagram matching.

Bundled files

The model reads these on demand while the skill is loaded. They are exposed as readable files and are never executed.

Frequently asked questions

What does the Visio Image Rebuilder AI skill do?

Rebuild or restyle editable Microsoft Visio diagrams from reference images and existing .vsdx files, then export deliverables as .vsdx, PNG, SVG, PDF, or PPTX. Use when the user asks Codex to open Visio, recreate a diagram from a PNG/JPG/screenshot/reference image, match a scientific model figure, update colors/typography/layout in a .vsdx, calibrate panel coordinates to avoid shifted or overlapping submodules, or make publication/presentation outputs while preserving Visio native editable shapes rather than embedding the reference image as a flat picture.

Why use Visio Image Rebuilder on TypingMind?

Because you install it once and use it with any model. Visio Image Rebuilder is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Visio Image Rebuilder in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/pengjunchi0/codex-visio-paper-figure-skill/tree/main. TypingMind reads its SKILL.md and bundles its files and installs it as a skill you can enable per chat.

Which AI models can use Visio Image Rebuilder?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Visio Image Rebuilder?

As many as you like. As long as a model supports skills, you can use Visio Image Rebuilder with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Visio Image Rebuilder AI skill free?

Yes. It is published on GitHub by pengjunchi0 under the MIT license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇