No user-submitted related servers found.
Scored across 19 tools
Core interaction tools (click, type, press_key, wait, navigate) are clearly distinct, but the read/extract cluster overlaps heavily: get_content, extract_data, see_page, and deep_analysis all retrieve page information, and an agent must read long descriptions to choose. Several tools (network_recorder, media_extractor, storage_inspector, extract_data) bundle many actions behind one name, adding internal selection ambiguity.
Most names are snake_case, but conventions are mixed: noun-first compounds (browser_init, browser_close), verb-first (get_content, execute_js), and bare verbs (wait, navigate, click, type). Individually readable, yet there is no single predictable pattern to infer names.
19 tools is on the heavy side for a browser automation server, and the multi-action mega-tools effectively expand the surface well beyond 19. Most tools are justified by real capabilities, but some inspection tools could be folded into see_page/extract_data.
The lifecycle is well covered: init/close, navigation and tabs, interaction, scrolling, waiting, content/media/network/storage extraction, captcha, redirect tracing, and async jobs. Minor gaps exist for dropdown selection, file upload, and dialog/alert handling, but agents can work around these with execute_js.