No user-submitted related servers found.
Scored across 35 tools
Most tools sit on distinct resources (memories, episodes, profiles, associations, persistence) with readable boundary docs — reconstruct vs recall vs get_contents vs recall_with_context are explicitly tiered, not redundant. Real overlap clusters remain around diagnostics (check_health vs deep_check vs get_session_findings) and listings (list_memories/list_episodes vs recall), where an agent could pick the wrong probe.
Almost everything is snake_case with a verb or noun prefix (get_/set_/list_/delete_/update_/lock_/pause_), so the pattern is predictable and readable. Minor deviation: several bare verbs (store, traverse, reconstruct) and noun-first names (persistence_status, check_health, get_operating_context) break the verb_noun rhythm, though not comprehension.
35 tools is well past the point where an agent can hold the surface in view; the domain (memory CRUD, recall tiers, persistence control, diagnostics, migrations, self-update) is broad but the count is heavy. At least one tool is already dead (open_dashboard, deprecated/removed in 2.7) and others (get_queue_status, check_update) are peripheral to the core purpose.
Coverage is strong: full memory lifecycle (store/recall/expand/update/lock/unlock/delete), episodes, profiles, associations with retract, bulk export/import/merge, persistence pausing, health/diagnostics, and precision tuning with read-back. Minor gaps: no update_episode (episodes are archive/delete only), no agent/registry listing tool, and a stale deprecated tool left in the surface.
