Bps Id

There are several overlapping query tools (ask_pipeworx, ask_pipeworx_beta, ask_pipeworx_grounded, deep_research, validate_claim) with nuanced differences that could cause misselection, and the polymarket suite has six tools with tightly related purposes (arbitrage, edges, edge tracker, fill risk, kalshi spread, bet research). These are distinguished by detailed descriptions, but the beta and stable router are currently identical, and the boundary between ask_pipeworx and deep_research requires careful reading. Most other tools (memory, subscriptions, entity resolution) are clearly distinct.

Nearly all tools follow a snake_case, verb-first or domain-prefixed pattern (ask_pipeworx, compare_entities, resolve_entity, polymarket_*). Some stray from the verb_noun ideal (e.g., ai_visibility_check starts with a noun, pipeworx_trending uses a domain prefix), but the overall style is consistent and readable. Exceptions like ask_pipeworx_beta are still within the same convention.

31 tools is on the heavier side but reasonable for a meta-server that aggregates thousands of underlying data sources across company filings, economics, prediction markets, and AI visibility. Each tool serves a distinct workflow (querying, deep research, arbitrage, memory, subscriptions, scanning), though some niche prediction-market tools (polymarket_edge_tracker, polymarket_fill_risk) could be seen as over-specialized for a general data platform. Still, the count feels justified for the breadth.

The surface is remarkably complete for its purpose: querying (ask_pipeworx), grounded verification (validate_claim, ask_pipeworx_grounded), entity resolution, comparison, company profiles, change feeds, AI visibility checks, memory persistence, subscriptions with alerts, and prediction-market analysis. Minor gaps exist (no tool for automated scheduling beyond subscriptions, no file management), but core workflows have no dead ends and the memory/subscription lifecycle is fully covered.