Scream Void

Most tools target distinct resources or workflows, but the set contains overlapping research, discovery, and prediction-market tools. In particular, ask_pipeworx_beta is explicitly identical to ask_pipeworx today, and ask_pipeworx, ask_pipeworx_grounded, deep_research, discover_tools, suggest_questions, and validate_claim form a cluster an agent must carefully distinguish. The long descriptions usually clarify preferred use, but overlaps still create real misselection risk.

All names use lower snake_case and several domain prefixes are consistent (polymarket_, updown_, pipeworx_), which makes the surface readable. However, the set mixes verb-first names (ask_pipeworx, validate_claim, resolve_entity) with noun-phrase names (company_facts, entity_profile, recent_changes, prediction_calibration), so there is no single predictable verb_noun pattern.

45 tools is well above the comfortable 3-15 range and past the 25+ 'too many' threshold. The breadth is partly justified for a universal data gateway, but the beta duplicate and the large specialized prediction-market suite make the surface heavier than an agent can reliably navigate.

For a broad data gateway, coverage is extensive: discovery/routing, grounded research, entity resolution, company facts, archives, prediction-market analytics, memory, and subscriptions all have dedicated tools. The main gaps are raw URL/resource fetching and subscription update, both workaroundable through existing tools.