| name | browser-test |
| description | Use this skill when the user wants to test, inspect, verify, or screenshot a page in the browser (e.g., "/browser-test", "/browser", "test in browser", "take screenshots", "verify in Chrome", "check devtools"). On Windows environments, this skill bypasses browser_subagent and uses Playwright MCP tools to navigate, evaluate JS/scroll, and capture screenshots reliably. |
Playwright Browser Testing & DevTools Verification
PURPOSE
Provides a bulletproof browser verification and screenshot workflow on Windows by leveraging Playwright MCP tools directly (ServerName: "playwright").
WHEN TO USE THIS
- When the user types
/browser-test, /browser, or requests browser testing on Windows.
- When you need to visually verify UI/UX changes, layout animations, scroll triggers, responsive behavior, or component renders.
- When
browser_subagent fails or is bypassed due to OS constraints.
EXECUTION INSTRUCTIONS
When triggered, execute testing using the following step-by-step Playwright MCP workflow:
-
Navigate:
Call call_mcp_tool with ServerName: "playwright" and ToolName: "browser_navigate":
{
"ServerName": "playwright",
"ToolName": "browser_navigate",
"Arguments": { "url": "http://localhost:3000/path" }
}
-
Evaluate / Interact:
Use browser_evaluate to execute JS (e.g. scroll to specific element, check offset, verify console errors):
{
"ServerName": "playwright",
"ToolName": "browser_evaluate",
"Arguments": { "function": "() => { window.scrollTo(0, 1000); return window.scrollY; }" }
}
-
Capture Screenshot:
Call browser_take_screenshot to capture visual evidence:
{
"ServerName": "playwright",
"ToolName": "browser_take_screenshot",
"Arguments": { "type": "png", "scale": "css" }
}
-
Diagnose & Report:
Inspect console outputs or errors returned by Playwright. Provide a visual verification summary report to the user with findings and verification state.