wwe-dog

wwe-dog / ppt-image2-editable-rebuild

Public

一个基于截图/参考图重建可编辑 PPT 的 Codex skill。先用 image2/imagegen 生成单页视觉参考图，再以局部截图保留复杂插图、图标、地图等视觉资产，同时用 PPT 文本框和形状重建标题、正文、卡片、箭头、标签、结论条等可编辑元素。

19
1
75% credibility
Found May 30, 2026 at 19 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Python
AI Summary

This is a productivity tool that helps you turn AI-generated PowerPoint mockups into actual editable presentations. Instead of just pasting static images into slides, it extracts complex graphics like icons and illustrations while rebuilding text, cards, and layouts as real PowerPoint elements. The result looks polished like the original mockup but remains fully editable—you can change words, adjust colors, or rearrange content anytime. It's designed for people creating academic presentations, business pitches, or any slides that need both visual appeal and flexibility.

How It Works

1
🎨 You have an idea for a beautiful presentation

You picture how your slides should look with stunning graphics, nice cards, and clean text layouts.

2
✨ You describe your vision and get perfect reference images

You tell the AI what each slide should look like, and it generates gorgeous visual mockups that capture your exact style.

3
🖼️ You pick the best parts from each reference image

You select which graphics, icons, maps, and illustrations to keep from your mockups, and which elements need to become editable text.

4
You work through your slides one by one
🖼️
Complex visuals stay as images

Fancy illustrations, icons, and decorative elements stay exactly as they look in your reference.

📝
Text and structure become editable

Titles, bullet points, cards, arrows, and labels turn into real PowerPoint elements you can change anytime.

5
👀 You check each slide against your reference

You preview your work to make sure nothing looks crooked, text isn't cut off, and everything matches your vision.

🎉 You have a perfectly editable presentation

Your slides look exactly like the beautiful mockups, but everything is fully editable, changeable, and ready to present.

Sign up to see the full architecture

4 more

Sign Up Free

Star Growth

See how this repo grew from 19 to 19 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is ppt-image2-editable-rebuild?

A Python Codex skill that converts AI-generated PPT reference images into fully editable PowerPoint files. It splits the workflow: image2/imagegen handles visual design, while the rebuild process reconstructs text, cards, arrows, and labels as editable PPT elements. Complex visuals like illustrations and maps are preserved via cropped screenshots rather than flattened into static images. The result is a presentation where you can actually edit the copy and restructure elements after generation.

Why is it gaining traction?

AI image generators create stunning PPT mockups, but the output is a static image—useless if you need to change a headline or adjust a timeline. This tool bridges that gap. Instead of dumping a full reference screenshot into your deck, it reconstructs each page as editable content while retaining visual fidelity for icons, maps, and badges. The process is systematic: generate per-slide references, crop assets, rebuild structure, review output. For developers building ppt generator workflows on Codex, this fills a real gap in the pipeline.

Who should use this?

Developers integrating ppt creation into AI agents or Codex workflows. Academic researchers preparing thesis defenses who want AI-designed visuals without sacrificing editability. Anyone building automated presentation pipelines who needs editable output rather than a gallery of PNGs. If you are generating PPT mockups with image2 and need the final deliverable to be a working document, this is purpose-built for that workflow.

Verdict

The concept is solid and the workflow is clearly thought through, but this is a small, early-stage project with 19 stars and a 0.75% credibility score. Documentation is mostly in Chinese and quality depends on how cleanly your image2 prompts generate reference slides. If you are already working in Codex and need this specific workflow, it is worth a try. For general audiences seeking a ppt creator, stick with more established tools until this gains community traction.

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.