No user-submitted related servers found.
Scored across 15 tools
Four tools retrieve memories (retrieveMemories, fetch_memories, searchMemoryIndex, startup_inject_memories) and three browse the store (listMemories, memoryTimeline, searchMemoryIndex), creating real selection risk. Descriptions do explain the distinctions (host-capability fallback vs. ranked search vs. index scan), but an agent must read carefully to avoid duplication, and startup_inject_memories overlaps heavily with retrieveMemories at session start.
Most tools use camelCase verb_noun (storeMemory, getMemory, forgetMemory, batchStoreMemory), but two use snake_case (fetch_memories, startup_inject_memories). Also, for a memory domain, the noun is sometimes 'Memory' and sometimes omitted (stats, handleSessionEnd), which weakens the pattern.
15 tools sits at the upper bound for a memory server, and several earn their place only in niche host configurations (fetch_memories, startup_inject_memories, resetAccessCounts). The core lifecycle could be covered with roughly 10 tools; the extra fallback/administrative surface makes the set feel heavy.
Coverage is broad: create, batch create, semantic retrieve, index search, timeline browse, get, list, delete, session events, summaries, and stats. The notable gap is updating/editing an existing memory's content or importance — only summaries upsert, so corrections require forget + store.