spacezephyr/space-multi-design-ppt — explained in plain English
Analysis updated 2026-05-18
Turn a written business plan or product pitch into a polished slide deck styled after a brand like Stripe or Linear.
Convert a weekly report into presentation slides using a specific brand's colors, fonts, and layout choices.
Get five recommended brand styles automatically when you are not sure which design language fits your content.
Export the same content as an editable PowerPoint file for further editing or as a PDF for sending and printing.
| spacezephyr/space-multi-design-ppt | amap-ml/blockpilot | ernie-research/nava | |
|---|---|---|---|
| Stars | 62 | 62 | 62 |
| Language | Python | Python | Python |
| Setup difficulty | moderate | moderate | hard |
| Complexity | 2/5 | 3/5 | 5/5 |
| Audience | pm founder | researcher | researcher |
Figures from each repo's GitHub metadata at analysis time.
Requires Node.js for design token fetching, and separately python-pptx/Pillow for PowerPoint output or Playwright/Chromium for PDF export.
This project is a skill for AI coding assistants like Claude Code, Codex, and Cursor that turns written content into a presentation slide deck styled after a real brand's design language. You give it an article, a business plan, a weekly report, a product pitch, or course material, and it produces slides that look like they were designed with the visual language of a company such as Apple, Notion, Claude, Stripe, Linear, or Tesla. Rather than guessing at colors and fonts, the skill pulls real design tokens (colors, fonts, letter spacing, corner radius, shadows) from a companion project called brand-design-md and a site called getdesign.md, so the visual choices trace back to an actual published design specification. If you do not pick a style, it looks at your content and recommends five matching brand styles plus a sixth smart match option. In total it supports 62 brand design styles, spanning tech companies, finance apps, and car and consumer brands. The process runs in five steps: understanding the content and its audience, deciding on a style, fetching the real design tokens for that style, generating the slide pages, and a final check for text overflow or layout problems before delivery. Output can be a single self contained HTML slide deck with page turning and a grid overview, an editable PowerPoint file built with python-pptx, or a PDF exported from the HTML version for printing or archiving. Installing it means either asking a compatible AI agent to install the skill directly from its GitHub URL, or cloning the repository into the agent's skills folder yourself. It needs Node.js to fetch design specs, and if you want PowerPoint output you also need python-pptx and Pillow, while PDF export needs Playwright and Chromium. The README notes the brand styles are design language inspired by public DESIGN.md tokens, not official brand templates, and that HTML output looks closer to the original brand than the PowerPoint version.
An AI agent skill that turns your article or report into a slide deck styled like a real brand such as Apple, Notion, or Tesla, using actual published design tokens.
Mainly Python. The stack also includes Python, python-pptx, Playwright.
No license information is given in the README.
Setup difficulty is rated moderate, with roughly 30min to a first successful run.
Mainly pm founder.
This repo across BitVibe Labs
Verify against the repo before relying on details.