Batchdata

Most tools are clearly distinct, but several overlap: ask_pipeworx_beta is explicitly identical to ask_pipeworx right now, polymarket_edges/polymarket_arbitrage/bet_research all probe prediction-market opportunities, and discover_tools/suggest_questions both serve as discovery entry points. The detailed descriptions help, but an agent could easily route a query to the wrong variant.

All tool names are snake_case and follow predictable product prefixes (ask_pipeworx*, polymarket_*, batchdata_*, pipeworx_*), which gives strong structural consistency. However, the pattern mixes verb-led names (remember, subscribe, validate_claim) with noun-phrase names (entity_profile, recent_changes, resolution_audit), so it is not a uniform verb_noun scheme.

At 39 tools this server is well over the 25-tool threshold for being considered too heavy. The breadth of the underlying data platform partially justifies it, but many tools are thin wrappers or variants (three ask_pipeworx flavors, three polymarket edge tools, separate remember/recall/forget primitives) that could be consolidated.

The server covers a broad domain well: universal lookup via ask_pipeworx, deep research, entity profiles, comparisons, property data, prediction markets, subscriptions, and memory. The universal router and discover_tools fill most gaps, so missing operations are minor (e.g., no dedicated news search tool, but ask_pipeworx handles it via routing).