Back to directory

Dashi PPT Skill

An AI-agent skill that generates browser-editable presentations across 12 visual themes and more than 1000 layouts, letting a user tweak layout, color and content live before exporting to PPTX, PDF or HTML.

Dashi PPT Skill

Dashi PPT Skill is an agent skill that turns a document into a presentation with 12 built-in visual themes, 1,020 layout pages and more than 8,500 adjustable controls. Each page renders as a real webpage with its own on-page control console, so a user can adjust it directly in the browser before exporting a real, editable PPTX file.

Layouts span about 20 page roles, such as cover, table of contents, metrics, trends, comparison and process pages, and include chart and analysis-model layouts such as radar, waterfall, treemap, funnel, heatmap, Sankey and Gantt charts, plus frameworks such as SWOT, Porter's Five Forces and the business model canvas. On-page controls let a user change layout, module count, color scheme and page emphasis; any text can be edited in place, and media slots can be replaced by click or drag.

Output exports as an offline HTML package, a PDF, or an editable PPTX that keeps its text editable. It works with several agent CLIs, including Claude Code and Codex, requires Node.js 20 or later, and needs a local Chrome, Chromium or Edge browser for PDF and PPTX export. It is written in JavaScript 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.