WebMCP

Use this skill when the goal is browser-native tool registration and direct in-browser tool execution, not just external browser control.

Use When

• integrating WebMCP into a dashboard, app, or workflow surface
• proving a page is WebMCP-ready in a real Chrome session
• debugging why tools register but cannot be directly executed
• validating browser-native operation before treating the page as an operator surface
• designing tool contracts, readiness gates, or troubleshooting paths for WebMCP-backed apps

Do Not Use When

• the task only needs DOM control, screenshots, console logs, or network inspection
• the page does not need browser-native tools
• the task is headless-only
• the page should mutate durable state directly instead of routing through audited backend actions