Back to directory

Guizang PPT Skill

An AI-agent skill for Claude Code and Codex that generates single-file HTML slide decks in an editorial-magazine or Swiss-design style, with matching cover images and a low-power presentation mode.

Guizang PPT Skill

Guizang PPT Skill is a skill for Claude Code, Codex and similar agent environments that generates single-file, horizontally paginated HTML slide decks, along with matching slide illustrations and cover images for several platforms. It ships two visual systems: an electronic-magazine and e-ink style suited to narrative and personal presentations, and a Swiss International style, grid-based with a single high-saturation accent color, right angles, hairlines and strong type-size contrast, suited to factual and methodology presentations.

Style A offers ten layout types such as cover, chapter, large-type data statements and image grids; Style B offers 22 fixed layouts such as cover, statement, KPI tower and comparison. An optional image pipeline can generate photos, infographics and diagrams sized to the deck's templates, and a built-in presenter mode gives a dual-window audience screen, speaker notes, timed rehearsal and automatic page turning; a low-power mode turns off WebGL and canvas animation.

The generated deck is a single HTML file that needs no build step or server. It installs with the skills command-line tool and is licensed under the AGPL-3.0 license.

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.