No user-submitted related servers found.
Scored across 80 tools
Most tools map cleanly to a single resource+action across well-separated domains (agents, calls, bulk calls, workflows, phone numbers, voice cloning), and descriptions explicitly disambiguate near-pairs like getWorkflow vs showWorkflow and createBulkCall vs startBulkCall. A few boundaries remain fuzzy: listAllProviders overlaps with listVoices/listLLMProviders/listSTTProviders/listTTSProviders, bulkCallActions (pause/resume/reschedule) sits ambiguously next to cancelBulkCall and startBulkCall, and addBulkCallContact vs addBulkCallContacts differ only by batch semantics.
Every name is camelCase verb_noun, so the surface is highly predictable overall. Deviations are minor but real: fetchBulkCalls breaks the list_* convention used everywhere else, bulkCallActions is a vague noun-first catch-all rather than a specific verb, and a handful of non-CRUD verbs (showWorkflow, canUploadFile, diffAgentVersion, dispatchCall) sit outside the dominant create/get/list/update/delete family.
80 tools is far past a comfortable agent-facing surface; even though the platform genuinely spans ten resource families, an agent must select among 80 similarly named endpoints per turn. Several clusters (bulk-call numbers/contacts/status/control, workflow node-type introspection in three separate tools) could be consolidated, and discoverability cost is high.
Coverage is strong where it matters: agents have full CRUD plus version history and per-action post-call management, workflows have a complete draft-validate-publish-activate-test lifecycle, and bulk calls, phone numbers, voice cloning and knowledge base all have lifecycle tools. Minor gaps remain: integrations expose only listIntegrations (no create/update/delete despite descriptions referencing creation), and sessions offer only createSession with no read/list counterpart.