zhongerxin/cowart

Cowart is a native infinite canvas widget plugin for Codex that provides visual drawing capabilities built on tldraw for ideation, annotation, image...

View on GitHub ↗Jump to charts ↓Open shareable report

Summary Information

Updated 22 minutes ago
Added to GitGenius on September 11th, 2026
Created on June 18th, 2026
Open Issues & Pull Requests: 22 (+0)
GitHub issues: Enabled
Number of forks: 445
Total Stargazers: 5,913 (+0)
Total Subscribers: 12 (+0)

Repository Insights (GitGenius)

Median issue/PR response: 20.6 hours
Mean response time: 40.7 hours
90th percentile: 3.3 days
Tracked items: 20

How this project is maintained

Around half of the issues opened in the past year never receive a reply. Only 23% of issues opened in the past year have been closed. Three people close 100% of everything that gets resolved.

Charts & Analytics

Fetching additional details & charts...

Issue Activity (beta)

Open issues: 11
New in 7 days: 1
Closed in 7 days: 0
Avg open age: 42 days
Stale 30+ days: 7
Stale 90+ days: 0

Recent activity

Opened in 7 days: 1
Closed in 7 days: 0
Comments in 7 days: 0
Events in 7 days: 0

Top labels

No label distribution available yet.

Most active issues this week

No issue events were indexed in the last 7 days.

Detailed Description

Cowart is a native infinite canvas widget plugin for Codex that provides visual drawing capabilities built on tldraw for ideation, annotation, image generation, and iterative image refinement.

The tool solves the problem of integrating a flexible visual workspace directly into Codex without requiring external browser windows or local web servers. It works by embedding a tldraw-based canvas as a native widget that opens within Codex itself. Users can create AI-powered image frames where they input prompts and reference images, and Codex generates pictures at the specified canvas position and dimensions. The tool also supports creating AI-generated HTML frames in 16:9 aspect ratio for executable single-file content, organizing these into AI Slides presentations with thumbnail previews and fullscreen playback, and annotating images to submit back to Codex for clean regeneration. Canvas data and resources persist to the current project's canvas directory rather than the plugin repository.

Cowart follows the Agent Plugins v1.0.0 specification with portable plugin entry points while maintaining Codex-specific metadata. The tool suits developers working within Codex who need to combine visual design, AI-assisted content generation, and iterative refinement in a single workspace. Installation is straightforward through Codex's marketplace system, and the plugin ships with prebuilt MCP bundles and widget artifacts, requiring no npm installation or network access at runtime. A complete Codex restart after installation is necessary for all skills and MCP tools to load fully.

The project's maintainers typically respond to new issues and pull requests within a day.