No user-submitted related servers found.
Scored across 33 tools
Each tool has a clear primary resource — memory, graph entity, procedure, decision, or account — and the detailed descriptions separate them well. A few boundaries could still trip up an agent (hebbrix_search vs hebbrix_ask vs hebbrix_graph_query; hebbrix_remember vs hebbrix_remember_many vs hebbrix_import), but these are not duplicates.
All tools share a clean hebbrix_ snake_case prefix and most use recognizable actions like create, get, list, update, delete, search, import, and export. However, verb/noun order is not uniform — e.g., claim_start vs account_status, graph_query vs search_entities — so the naming pattern is predictable only within clusters.
33 tools is well into the 'too many' range for a single MCP server and forces agents to hold a large tool surface in context. The tools are organized into clear subdomains, so the count is not chaotic, but several niche capabilities (claim workflow, extraction_status, learning_insights) could be consolidated or nested.
The memory lifecycle is complete — write, batch write, read, update, delete, search, history, export, and import — and knowledge-graph, procedure, decision-learning, and account surfaces are all represented. Minor gaps remain, such as no procedure-execution history/audit tool and no collection-level management beyond listing and exporting.