Back to directory

Guizang Social Card Skill

A Claude Code / Codex skill that turns an article, screenshot or video into Xiaohongshu image carousels and WeChat cover-image pairs, rendered from single-file HTML to PNG.

Guizang Social Card Skill

Guizang Social Card Skill is an agent skill for environments such as Claude Code and Codex that generates social media image cards from an article, marketing copy, screenshots, product notes, subtitles, photos or user video, producing Xiaohongshu (Rednote) image carousels, Live Photo animated cards, and WeChat 21:9-plus-1:1 cover pairs. It shares a workflow with a companion PPT skill but is maintained separately: the PPT skill handles horizontal slide decks, while this one handles static social image feeds.

It includes two built-in visual systems -- an Editorial style modeled on restrained magazine layouts for narrative or lifestyle content, and a Swiss International style using a grid, a single anchor color and stark type contrast for product reviews, data or tutorials -- across three canvas sizes and 28 layout templates with 10 preset themes. Live Photo cards support single, two-up, three-up and four-up video layouts at 3 or 5 second lengths depending on platform.

An image-sourcing workflow prioritizes user-supplied images before falling back through Unsplash, Pexels, Flickr Creative Commons, Wallhaven or direct search, and a validation script checks for overflow, font-size limits, density and footer collisions before the single-file HTML is rendered to PNG with Playwright.

More in Agent skills

See all 581
  • Provenance gate

    github.com

    A deterministic authorization gateway that sits between an AI agent and its tools, tracing where each parameter of a side-effect call came from, and blocking the call or asking a human when the source does not check out.

  • A markdown framework of system rules that restates Isaac Asimov's Four Laws of Robotics as a set of prioritized loss calculations for large-language-model agents to follow, published as a discovery file for agents and crawlers to read.

  • Vitruvius

    github.com

    An engineering research agent that runs a discover, read, synthesize, verify and review loop over technical questions, reading sources directly rather than inferring from titles, and publishing its own scored verifier benchmark, weaknesses included.