No user-submitted related servers found.
Scored across 13 tools
The four read paths (search_memory, search_keywords, list_memory, query_memory) overlap conceptually, but each description explicitly contrasts itself with the others (ranked vs. unranked, hybrid vs. BM25, structured MindQL) so an agent can pick correctly. Similarly, get_edges (edges touching one thought) vs. list_edges (global filtered edge browse) is distinguishable but confusable on first read.
Most names follow verb_noun, but conventions drift across entities: thoughts use store/get/update/delete_thought while edges use link_thoughts plus delete_edge, and memory_stats is the one noun-only name. The entity vocabulary also shifts between 'thought' and 'memory' (store_thought vs list_memory/search_memory), which muddies the pattern.
13 tools sit comfortably in the well-scoped range for a memory-graph server, with reads, writes, deletes, and analytics each represented. No tool looks like filler or a near-duplicate of another.
Thoughts have full create/read/update/delete plus ranked, keyword, structured, and paged listing, and edges have create/read/delete; deletion cascades are documented. The notable gap is that edges have no update operation, so weight or metadata on an existing edge cannot be changed without deleting and re-linking it.