Related Servers

There are many overlapping search/retrieval tools whose boundaries are hard to infer from names alone: quipu_search, quipu_search_nodes, quipu_search_facts, quipu_hybrid_search, quipu_unified_search, quipu_search_query, quipu_context, quipu_ask, and quipu_spotlight all sound like ways to find knowledge. Similar confusion exists between event tools (quipu_events vs quipu_graph_events, quipu_changes, quipu_transactions) and between ingestion paths (quipu_knot, quipu_set, quipu_update, quipu_episode, quipu_episodes_complete). Descriptions are detailed, but the set as a whole has multiple unclear boundaries.

The quipu_ prefix and snake_case style are consistent throughout. However, naming conventions are mixed: many tools are verb_noun (quipu_graph_create, quipu_merge_apply), while others are noun-only or domain metaphors (quipu_entity, quipu_graph, quipu_context, quipu_cord, quipu_knot, quipu_unravel, quipu_ask). The pattern is readable but not predictable enough for a strong score.

Eighty-four tools is far beyond a manageable surface for an agent, even for a feature-rich knowledge graph server. The set includes many niche governance, overlay, import/export, and event-cursor tools that will be hard to discover and select correctly. This is an extreme mismatch between count and practical usability.

The tool surface covers an unusually broad lifecycle: entity minting, fact assertion/retraction, SPARQL query/update, semantic and keyword search, SHACL validation, schema proposals, reasoning, events, imports/exports, overlays, graph freezing, and governance verdicts. No obvious core operation appears missing from the stated knowledge-graph and provenance domain.