← LightState.ai
Query LLM request logs, monitor caching hit rates, track per-user spend, and debug prompt failures via Helicone MCP.