No user-submitted related servers found.
Scored across 12 tools
Each tool targets a distinct resource plus action: notes (remember/search/node_get/node_update/recall), the task queue (submit/list/update), messaging (send/get/mark_read), and health. The only naturally confusable pair, brain_search vs brain_recall, is explicitly separated in the descriptions (ranked note list vs assembled context block within a token budget), and remember vs update is likewise clarified. No two tools appear to do the same thing.
All names are snake_case under a uniform `brain_` prefix, and most follow a resource-then-verb form (brain_tasks_submit, brain_messages_mark_read). Minor deviations exist: bare verbs (brain_remember, brain_search, brain_recall, brain_health) and a singular/plural resource split (brain_node_get vs brain_tasks_list). Readable and predictable overall, but not a single rigid pattern.
Twelve tools is well inside the healthy range and each one maps to a real operation across the three sub-domains (memory, task queue, messaging). Nothing looks like filler, and the surface is small enough that an agent can enumerate it easily.
The memory lifecycle (create, read, update, search, recall) and the task-queue and inbox lifecycles are covered end to end, including dependency release on completion. Gaps remain: no way to delete or archive a note or message, and no single-task fetch (you must list and filter), though status changes via brain_tasks_update give a partial workaround.