Geo Reconcile

The set contains several near-duplicate families: ask_pipeworx / ask_pipeworx_beta / ask_pipeworx_grounded are three versions of the same universal router, and the Polymarket tools (edges, arbitrage, kalshi_spread, fill_risk, edge_tracker, bet_research) plus company tools (company_facts, entity_profile, compare_entities, recent_changes, resolve_entity) share heavily overlapping input domains. Descriptions work hard to disambiguate with explicit 'PREFER OVER' and initialism guidance, which keeps it usable, but an agent still faces real selection risk across these clusters.

All names are lowercase snake_case, which is consistent throughout. However the convention is mixed: verb_noun forms (resolve_entity, validate_claim, compare_entities, generate_llms_txt) sit alongside bare nouns (company_facts, entity_profile, deep_research, recent_alerts) and bare verbs (remember, recall, forget, subscribe). Readable and predictable within each family, but no single unified pattern.

37 tools is heavy for any server, and the nominal scope ('Geo Reconcile') is covered by exactly one tool (reconcile_wikidata_osm), so the count is wildly disproportionate to the stated purpose. Many of the extra tools are meta-routers or overlapping variants that inflate the surface without adding distinct capability.

As a broad Pipeworx data-aggregation surface it is fairly complete: company data, macro, prediction markets, memory, and monitoring subscriptions all have read and some lifecycle coverage. But the apparent geo-reconciliation domain has only a single relevant tool, and no coverage exists for common siblings like bulk OSM/Wikidata editing or other geo joins, leaving the stated purpose thin while unrelated domains are expansive.