Audit voice-agent RAG grounding citations
Aggregate how grounded your voice agents are over a bounded window: totals (voice turns, turns with retrieval, cited vs uncited retrievals, citation coverage rate, no-retrieval share, low-confidence cross-references), a per-agent breakdown, the most-recent uncited retrieval turns as a triage list, and a daily coverage trend. Voice-scope only — chat turns are excluded; unscoped_turns counts turns whose conversation row was unreadable so exclusions stay visible. Owner or admin role. Query: agentId, from/to (ISO-8601), days (window fallback), lowConfidenceThreshold, turnLimit.
Authorizations
Dashboard JWT token from Clerk
Headers
Sandbox opt-in for Clerk-session-authenticated requests. Set to true to route the call through the test-mode pipeline: no real provider delivery, no credits deducted, response meta.test_mode: true. Ignored for live API keys (dv_live_sk_*) — server-to-server clients must use a test-prefixed key (dv_test_sk_*) to exercise sandbox. Test-prefixed keys unconditionally enable sandbox regardless of this header.
true, false Query Parameters
Narrow to one agent.
1 <= x <= 365