Umbraco Backoffice Navigation (Claude-in-Chrome)

Why this skill exists

The Umbraco backoffice is built from Lit web-components. That creates three problems for browser automation that will waste many tool calls if you fight them blind:

1. Everything is in (open) shadow DOM. The built-in find and read_page tools do not pierce shadow roots, so they return almost nothing on backoffice pages.
2. Modals portal into their own stacking context and frequently render off the ~1536px captured viewport, or stack so the newest one is off-screen.
3. Screenshots lag behind the live DOM during modal transitions — you can screenshot a frame that no longer reflects the current state and misread it completely.

The fix is to stop navigating by pixels. Inject a small shadow-DOM-piercing helper and read state from the DOM (which never lags), acting by label/text instead of coordinates. Take a screenshot only when a picture is the actual deliverable (verifying how something looks, or evidence for the user) — not as your primary way to figure out what's on screen.