Scored across 27 tools
Each tool targets a distinct resource and action: discovery (apps, windows, find, snapshot, screenshot), element inspection (get_element), input (click, set_value, set_numeric_value, focus, select_text, type_text, press_key), and window management (activate, minimize, close, move, resize, fullscreen). Overlaps like find vs wait_for are differentiated by polling behavior, and mouse_move/scroll have distinct roles.
All tools follow a consistent lower_snake_case verb_noun pattern (mouse_move, minimize_window, activate_window, set_value, wait_for_window, etc.). Exceptionally uniform, with no mixed conventions or vague verbs. Even atypical names like 'find', 'snapshot', and 'action' function as clear verbs.
At 27 tools, the count exceeds the typical 3-15 range, but the domain (cross-platform UI automation) justifies the breadth. Each tool covers a distinct operation—discovery, inspection, input, window manipulation, and waiting—with no redundancy. The count feels justified by the complexity required to handle various platforms and web content.
The surface is remarkably complete for UI automation: it covers discovery (apps/windows/elements), inspection (snapshot, get_element, read_text), interaction (click, type, set_value, scroll, drag via coordinate/action), and synchronization (wait_for, wait_for_app, wait_for_window). The generic `action` tool covers any raw accessibility action, providing a fallback for unforeseen operations. No critical dead-ends.