My AI QA agent said "all features working." The canvas was blank. Here's what it was actually seeing.

Back when I first delegated QA to an AI agent, it signed off on a tool with "all features working, pass." I opened the tool myself. The canvas was blank.

The AI wasn't lying. In the environment it was looking at, the tool genuinely appeared to work.

I run visual QA across a large fleet of web tools using Claude + Chrome MCP, and this class of false positive traced back to exactly two causes.

Cause 1: requestAnimationFrame stops in hidden tabs

Chrome MCP typically operates in hidden (background) tabs. And browsers throttle requestAnimationFrame aggressively in hidden tabs to save power.