No user-submitted related servers found.
Scored across 32 tools
Most tools target distinct resources (caches, queries, metadata, connections, semantic tables) and the descriptions clearly separate them. The main ambiguities are peaka_execute_sql_query vs peaka_execute_query (raw SQL vs saved query by ID) and peaka_query_golden_sqls, which sounds like execution but is actually retrieval. Granular discovery tools overlap somewhat with peaka_get_project_metadata, but descriptions clarify their scoped roles.
All tools share the peaka_ prefix and consistently use snake_case verb_noun naming, which is predictable. Minor inconsistencies exist: plural status tools use get_ (get_cache_statuses, get_materialized_query_statuses) instead of list_, and query is used as a verb in peaka_query_golden_sqls. Overall the create/update/delete/refresh/list/get conventions are followed well.
32 tools is heavy for an MCP server, falling in the too-many range. The set spans several subdomains, but many granular metadata tools (list_catalogs, list_schemas, list_tables, list_columns, plus get_project_metadata) could be consolidated. Even though each tool has a purpose, the overall surface feels bloated.
Core workflows are well covered: saved queries have full CRUD plus execution and refresh, caches have create/update/delete/refresh/status, and metadata discovery is thorough. The main gaps are in the semantic layer—no delete_semantic_catalog or update_semantic_table—and connections are read-only. These are workable gaps rather than fatal dead ends.