No user-submitted related servers found.
Scored across 45 tools
Most tools target distinct resource+action pairs, and the descriptions actively resolve the close calls (dashboard-vs-raw for inventory/CVE listing, and 'relevant' vs global catalog for the three planning pairs). A few pairs remain subtly overlapping, such as advisor get_hosts_hitting_a_rule vs get_hosts_details_for_rule and rbac get_caller_access vs get_caller_access_all, but explicit 'prefer X over Y' guidance keeps misselection unlikely.
Names follow a predictable namespace__verb_noun snake_case pattern across almost all tools (inventory__list_hosts, vulnerability__get_cves, advisor__get_active_rules). Minor deviations exist: get_mcp_version has no namespace, 'load' verbs are mixed with 'get' for the two dashboard tools, and namespaces use hyphens (image-builder, content-sources) while others use underscores.
45 tools is heavy for a single agent surface and well above the 25-tool threshold, making discovery and selection costly. The count is partly defensible because it spans nine distinct Red Hat domains (inventory, advisor, vulnerability, planning, image-builder, rbac, rhsm, content-sources), but as one server it remains oversized.
Coverage is broad and coherent for a mostly advisory/read platform: inventory, vulnerability, advisor, planning lifecycle, image-builder, RHSM, and RBAC each have their core read operations. Gaps are read-only in nature — image-builder has no create-blueprint/compose tool and RHSM only lists activation keys — which agents can work around but limits true lifecycle coverage.