Scored across 63 tools
Most tools are cleanly separated by vertical plus action (e.g. Amazon search vs product vs reviews vs seller), so an agent can usually tell them apart. However there are genuine collision zones: two tools both named getMapReviews (place reviews vs contributor reviews), a full Google SERP and a 'light' variant that differ only by cost/layout, and several overlapping place-detail tools across Google Maps, Yelp, and YellowPages. These clusters create real misselection risk, but the descriptions do disambiguate in most cases.
The dominant pattern is hasdata_<vertical>_<sub>_<action>, which is mostly followed. But verbs are inconsistent (get, perform, scrape, search) — e.g. performMapSearch and scrapeWebPage break the get_ convention — and one tool carries a junk hash suffix (getImmersive_e29f691177). Readable overall, but not a clean predictable scheme.
63 tools is very heavy by any standard, well beyond the 25+ threshold that signals overload. It is partly justified because the server aggregates ~30 distinct scraping verticals with search/detail/reviews each, but the sheer breadth makes discovery and selection costly for an agent.
The surface is broadly complete for a scraping aggregator: most verticals ship a search/topic tool plus a detail tool plus a reviews or secondary-detail companion (Amazon, Walmart, Yelp, Google Maps, Airbnb, Indeed, Zillow/Redfin, YouTube). Minor gaps exist (e.g. no explicit pagination-agnostic batching or cross-vertical aggregation), but core lifecycles are covered.
