open-slide/open-slide

What it solves

Creating presentation decks often involves tedious manual layout work. open-slide solves this by providing a framework specifically designed for AI coding agents to generate, edit, and manage slide decks using natural language prompts, turning the slide-creation process into a code-driven workflow.

How it works

open-slide provides a React-based runtime that renders slides onto a fixed 1920x1080 canvas. You initialize a project using the CLI, which includes built-in "agent skills" (system prompts) that guide AI models like Claude or Cursor on how to structure decks, manage layouts, and apply styling. The framework includes an in-browser inspector that allows you to leave visual comments on slides; these comments are saved as code markers, which the AI agent can then read and automatically apply to the source code.

Who it’s for

This is for developers and power users who prefer writing code or using AI agents to build presentations rather than using traditional drag-and-drop slide software. It is ideal for those who want to maintain version control over their decks and automate the design process.

Highlights

• Agent-native authoring: Includes specialized prompts and skills for AI agents to draft entire decks from a single prompt.
• In-browser inspector: Allows you to pin comments directly onto slide elements, which the AI agent can read and resolve automatically.
• Professional present mode: Features a built-in presenter view with speaker notes, slide previews, and a timer.
• Versatile exports: Supports exporting decks to static HTML, print-ready PDFs, and editable PowerPoint (PPTX) files directly in the browser.
• Asset integration: Includes an integrated search for brand logos via the svgl catalogue and a centralized asset manager.