agent-loop-browser

name	agent-loop-browser
description	Browser automation and computer-use skill for the agent-loop plugin. Takes screenshots, validates UI behavior, fills forms, clicks elements, and performs live web proof. Uses existing Chrome session via DevTools MCP.

Agent Loop Browser

Browser automation for the agent-loop. Use this for:

• Taking screenshots of deployed previews for visual validation
• Filling forms, clicking buttons, navigating pages
• Checking live UI behavior as part of PR verification
• Capturing before/after proof for autonomous PRs

Prerequisites

Prefer the existing Chrome session (logged-in state, extensions, cookies).

DevTools MCP Check

# Check if Chrome DevTools MCP is available
mcporter list chrome-devtools --schema 2>/dev/null || echo "MCP not available"

If MCP is unavailable, fall back to the browser skill's CLI:

# Check browser CLI
command -v browse >/dev/null 2>&1 || echo "browse CLI not installed"

Page Navigation

# Navigate to URL
mcporter call chrome-devtools.navigate_page \
  --args '{"url":"https://github.com/owner/repo/pull/123"}' --output text

# Or with browse CLI
browse goto "https://github.com/owner/repo/pull/123"

Taking Snapshots

# Text snapshot of current page (for understanding layout)
mcporter call chrome-devtools.take_snapshot --args  --output text


mcporter call chrome-devtools.take_screenshot \
  --args