Paper-Presentation-Tips Adds AI Skill for Generating & Polishing Academic PPTs

The Paper-Presentation-Tips project organizes 11 checkable suggestions for academic paper presentations across five categories: overall design, background motivation, method explanation, experiment analysis, and closing summary. Each tip follows a “Common Issue, ❌ Bad Slide, ✅ Good Slide, 💡 Core Principle” format with before-and-after slide comparisons. Six representative tips address common problems:

Too much text on a slide, key points buried → Tip 1: Keep word count per slide low

One slide tries to cover background, method, and results → Tip 2: One slide, one core task

Long background but audience still misses the problem → Tip 6: Use a concrete example to introduce the problem

Formulas appear too early, audience loses track → Tip 7: Show an example first, then the formula

Many experimental results, core finding not highlighted → Tip 9: Let the audience see the key point at a glance

Abrupt ending without a clear recap → Tip 11: Review core content on the summary slide

These tips can be used when preparing a new talk or auditing an existing deck to refine content selection, narrative order, and visual emphasis.

New paper-presentation-skill : AI-Assisted PPT Generation & Polish

The skill embeds the presentation advice into a three-step production pipeline that outputs an editable .pptx file:

Outline — Provide the paper, audience background, and time limit; the agent structures the narrative flow and per-slide content. For an existing PPT, it first lists concrete issues (why they hinder understanding) and proposes specific fixes.

Design — After confirming the outline or revision plan, the agent renders full-page reference images so you can review text, main figures, color scheme, and layout slide by slide. You can request adjustments such as “slide 3 has too much information” or “the conclusion on slide 7 needs to be larger.”

Build — The agent creates the confirmed design as shapes, tables, and charts in a .pptx file, then exports page images for a final check of small text, overlaps, and layout issues. Photographs and complex original paper figures are kept as images so they remain movable, scalable, and croppable.

The repository includes before-and-after examples such as “using input-output cases to illustrate a task” and “unifying color scheme and font size.”

Installation

Prerequisites: Git, Python 3.10+, and an agent that supports skills (e.g., Codex or Claude Code) with file read/write, command execution, and image viewing capabilities. Run:

git clone https://github.com/MLNLP-World/Paper-Presentation-Tips.git
cd Paper-Presentation-Tips
python -m pip install -r skills/paper-presentation-skill/requirements.txt

Final output is verified in PowerPoint, WPS, or LibreOffice; reference images can be produced by the agent’s built-in drawing tools, a configured drawing service, or user-supplied images. Detailed configuration and alternative installation steps are in the repository’s usage guide.

Usage Examples

Create a new presentation from a paper:

/paper-presentation-skill
请根据提供的论文制作一份中文组会 PPT。
听众：本领域研究生。
时长：约 15 分钟。
重点：讲清研究问题、方法核心思路和主要实验结论。

请先整理汇报大纲与每页内容，等我确认后再展示页面参考图；参考图确认后，制作可编辑的 PPTX，并检查实际显示效果。

Polish an existing PPT:

/paper-presentation-skill

请检查并润色这份 PPT。
结合仓库中的汇报建议，逐页指出内容安排、信息密度、标题、配色和图表中的具体问题，并说明怎么改。

先给出问题清单和修改方案，等确认后再生成页面参考图、制作可编辑的 PPTX。

This stepwise confirmation lets you lock down content and design before checking the final rendered .pptx.

Project Resources

Beyond the 11 tips and the AI skill, the repository collects research-presentation videos, courses, experience articles, and related tools. Contributions via Issue or PR are welcome — share presentation experiences, add Bad/Good Slide cases, or pass the project to peers preparing group meetings, conference talks, or academic showcases.

GitHub repository: https://github.com/MLNLP-World/Paper-Presentation-Tips