Arcgis Yakima

Several tools overlap heavily: ask_pipeworx, ask_pipeworx_beta, ask_pipeworx_grounded, and deep_research all route questions to the same universe of data sources, differing mainly in mode (beta, grounded, multi-hop). The polymarket_* family (edges, arbitrage, fill_risk, edge_tracker, bet_research) also has overlapping purposes around finding and evaluating prediction-market opportunities. An agent could easily misselect among these.

Tool names mix conventions: some are clear verb_noun (query_layer, search_datasets, resolve_entity, scan_dependency), while others are brand-prefixed (ask_pipeworx, pipeworx_feedback, polymarket_edges) or generic verbs (remember, recall, forget, subscribe, unsubscribe). The pattern is not uniform, but most names are readable and indicative of their function.

The server hosts 35 tools, which is at the high end of the appropriate range. Many are meta-tools that fan out to thousands of underlying packs, so the count is defensible, but it feels slightly heavy for an agent to navigate without strong discovery support. A few tools (ask_pipeworx_beta, scan_competitor_ai_presence) are marginal and could be merged or removed.

The tool set covers a vast data domain (SEC, economics, prediction markets, geospatial) but has notable gaps: there is no update or delete for many resources (e.g., no way to modify a subscription beyond cancel, no CRUD for stored memories beyond save/recall/delete), and the geospatial side is thin (only search/query/layer_info, no editing or analysis). The set is broad but shallow in lifecycle coverage, and some operations are dead ends.