No user-submitted related servers found.
Scored across 15 tools
Multiple tools cover retrieval/reading of the same underlying corpus: rsi_recall, rsi_knowledge_search, rsi_explore and rsi_memory all pull knowledge/memory content, distinguished only by entry path. Names like rsi_review vs rsi_knowledge_review, and rsi_stats vs rsi_impact vs rsi_memory_health, look interchangeable and require reading long descriptions to separate. Several tools are also multi-action multiplexers (rsi_learn, rsi_review, rsi_memory, rsi_conflicts) whose internal actions overlap with other tools.
The rsi_ prefix and snake_case are used throughout, which is good. However, the verb/noun convention is mixed: some tools are bare nouns (rsi_memory, rsi_stats, rsi_impact, rsi_conflicts) while others are verb or noun_verb (rsi_knowledge_search, rsi_knowledge_add, rsi_learn, rsi_recall). The near-identical rsi_review / rsi_knowledge_review pair further weakens predictability.
15 tools sits at the upper bound of a healthy set for a fairly broad surface (recall, knowledge CRUD, review/approval, conflicts, health, stats/impact, learning packs). Each tool broadly maps to a distinct capability area, though a few analytics/health tools feel like they could be consolidated. Not an egregious over-scoping.
Knowledge lifecycle is well covered: add, update/merge, delete, search, browse, approval review, conflict resolution, feedback, and health remediation. The main gap is a direct read-by-id/get-item tool; reading full bodies relies on rsi_recall semantics, which is a minor workaround rather than a blocker.