Once an agent is live, the questions change: why did it loop nine times, which step used the tokens, what did it send. AgentOps records sessions so you can replay them instead of guessing from logs.
Reach for it when
- You have an agent in production and cannot see inside it
- Token spend is surprising you
- You need to replay a failed run exactly
Look elsewhere when
- You are still prototyping locally
- Your framework's built in tracing is enough
What the lessons will cover
- 01Where AgentOps sends data
- 02A local collector
- 03Span kinds
- 04What inputs are recorded
- 05LLM calls
- 06Traces and tags
- 07Errors and guardrails
- 08A traced desk with privacy tests
Others in observability
Common questions
Is AgentOps free to use?
AgentOps is open source and free to run yourself. You still pay whichever model provider you point it at, and this tutorial is free with no signup.
Do I need to know Python to use AgentOps?
Basic Python is enough. Functions, dictionaries and imports cover most of what AgentOps asks of you.
When should I not use AgentOps?
You are still prototyping locally. Your framework's built in tracing is enough.
Lessons for AgentOps are being written. Meanwhile the Agent Governance Toolkit tutorial covers the same ground: state, tools, loops, memory and human approval. Most of it carries straight over.
Start the Agent Governance Toolkit tutorial →