Back to directory

Cowart

A native infinite-canvas widget plugin for Codex, built on tldraw, for sketching ideas, annotating images, and generating and iterating images, HTML, and slide decks with Codex.

Cowart

Cowart is a native infinite-canvas widget plugin for Codex. Built on tldraw, it opens directly inside Codex through an MCP widget rather than a browser, and it persists its canvas pages and image assets in the current project's own folder rather than inside the plugin's repository.

On the canvas, a user can create AI image frames that generate or replace an image from a prompt and reference images at the frame's position and aspect ratio, create widescreen AI HTML frames that generate a runnable single-file HTML page from a prompt and embed it for further editing, and combine images and HTML into AI Slides presentations with thumbnail previews and full-screen playback. Annotated screenshots can also be submitted so Codex generates a cleaned-up new image next to the original, and a bundled MCP tool lets Codex read the current selection, save the canvas, or insert images and HTML into it.

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.