Agent Trace
A full record of everything your agent did in one run: every LLM call, tool invocation, reasoning step, and state change. Essential for debugging and observability.
Also calledexecution traceagent traces
Where you meet this in practice
Foundation of AI opens one full lesson to anyone. No account, no card, the whole lesson.
Read the free lessonAbout this glossary
Plain-English definitions of the AI and AI agent terms that turn up in vendor pitches, product decisions and team conversations.