Canvas Browser Inspector

•
7
•

Inspects and interacts with HTML canvas and WebGL web pages when standard DOM and accessibility tools fail.

Bridges the gap between Chrome MCP browser navigation and computer-use visual controls to inspect, read, and interact with web pages rendered entirely on HTML5 canvas or WebGL elements. When traditional DOM extraction, accessibility trees, and screen scrapers return empty output, this skill provides a battle-tested workflow combining full-screen visual captures, region-targeted zooming, pixel coordinate clicks, and hybrid JavaScript fallbacks. It enables automated interaction with modern graphics-heavy dashboards, charting suites, game-like UIs, and virtualized browser views.

Key Features

01Hybrid DOM traversal for targeting underlying HTML elements and event triggers
02High-resolution regional zooming for reading fine canvas text, tables, and charts
037 GitHub stars
04Precise coordinate-based clicking and interaction mapping for canvas-based elements
05Direct network data scraping workarounds using in-browser fetch execution
06Coordinated workflow bridging Chrome MCP navigation with computer-use visual capture

Use Cases

01Automating interactions and end-to-end testing on custom canvas UI engines that lack DOM nodes
02Scraping financial dashboards, trading charts, or analytics portals rendered via WebGL or HTML5 canvas
03Extracting tabular text and values from virtualized web applications when accessibility trees return empty