Glossary
Agent Observability
Instrumentation that reconstructs an agent's full execution path — tool calls, memory reads and writes, reasoning steps, and handoffs between agents — rather than just input, output, latency, and cost the way standard LLM-call observability does.
Standard LLM observability captures a single model call. Agent observability has to reconstruct a full execution path across multiple tool calls and handoffs — a different problem, not a bigger version of the same one.
Isn't agent observability just LLM observability with extra steps?
No — it's a different problem, not a bigger version of the same one. Standard LLM observability captures a single call's input, output, latency, and cost. Agent observability has to reconstruct a full execution path across multiple tool calls, memory reads and writes, and handoffs between agents, which needs different instrumentation entirely.
Where this comes up
AI Agent Observability: What to Track
Need more than a definition?
If this term describes a problem you have, find out where you stand with a free scored diagnostic.