Turn a research paper or thesis into a structured set of editable PowerPoint slides ready for a group meeting or journal club presentation.
Generate slide layout blueprints and image mockups for approval before producing the final PPTX file.
Use a reference template to keep the visual style consistent with Chinese academic presentation conventions.
Convert figures, screenshots, or hand-drawn mockups into properly laid-out presentation slides using Codex or ChatGPT.
Requires Codex or ChatGPT with skill-loading support, PPTX creation works as a local file workflow in Codex and as an export step in ChatGPT.
scholar-ppt-cn is a skill for Codex and ChatGPT that turns academic source material into editable PowerPoint presentations. It is designed for Chinese academic presentations, such as group meeting slides for journal clubs or research progress reports. The README is partly in Chinese and partly in English. The skill works through a structured multi-step process. First, it generates a production planning table that maps each planned slide to a narrative section, a source asset (like a figure or screenshot from a paper), the visual layout to use, and the core message that slide should communicate. That table then drives the creation of a mockup family, which is a set of slide layout blueprints with variations. From those blueprints, the skill either uses an image generation model to produce sample pages for approval, or generates an editable PPTX file directly. A final quality review step checks the output before the file is delivered. The skill accepts a range of input types: research papers, theses, reports, figures, notes, reference templates, screenshots, or hand-drawn mockups. The repository includes a five-slide reference PowerPoint template that establishes the visual style for Chinese academic presentations. In Codex environments, the skill treats PPTX creation as a local file workflow, meaning it generates files that can be inspected or previewed locally. In ChatGPT environments, the planning and blueprint stages work the same way, but the final local QA steps are done after exporting the file. Version 3.3.1, the most recent described in the README, added Codex-specific local execution rules, explicit reference template loading, and batched image generation for slide mockups.
← deathcats4 on gitmyhub — every repo by this author, as a profile.
Verify against the repo before relying on details.