No user-submitted related servers found.
Scored across 106 tools
Many tools overlap substantially: browser_page_text and browser_page_text_by_tab_id are explicitly documented as doing 'the same thing now', while browser_type, browser_fill, browser_set_editable_content, and browser_form_fill are four paths to entering text with subtle boundaries. Descriptions do help distinguish most cases (e.g. click vs trusted_click vs click_and_wait_navigation), but the sheer surface creates frequent ambiguity.
Nearly all tools use the browser_ prefix in snake_case, and verb_noun forms like browser_navigate, browser_close_tab, and browser_set_policy are predictable. However, a large minority are bare nouns with no verb (browser_health, browser_logs, browser_cookies, browser_forms, browser_html, browser_markdown), which slightly breaks the pattern.
106 tools is far beyond what any single agent can scan effectively, well past the 25+ threshold for a heavily overloaded surface. Many of these exist as niche one-offs (browser_wp_html, browser_reload_extension, browser_broadcast) that dilute discoverability of the core set.
The surface covers browser automation exhaustively: navigation, tabs, windows, cookies, storage, sessions, network interception/capture, screenshots, forms, JS execution, waiting primitives, policy, tagging, pooling, and diagnostics. It is hard to identify a meaningful missing operation for the stated domain.