Tag: Tracing

Agent execution trace timeline showing a tool_call span next to the final narrated text, used to verify an AI agent actually called the tool

How to Verify an AI Agent Actually Called the Tool

Learn how to verify an AI agent actually called the tool: diff…

By Surya Koritala

31 Min Read

A Grafana Tempo trace waterfall showing an AI agent's invoke_agent parent span with nested chat and execute_tool child spans, self-hosted on a local LGTM stack

Self-Host LLM Observability With OpenTelemetry in 2026

Learn to self-host LLM observability with OpenTelemetry: a free Grafana LGTM stack,…

By Surya Koritala

34 Min Read