Skip to content

AI Observability

Traces of thoughts, tools, tokens and costs.

prompt loggingtracestool callstoken usagecostevaluation scores

Lessons

  1. 34.1Agent traces: seeing what the AI actually didWhen an AI agent does something strange, a trace shows you exactly what it saw, did and cost, step by step. 30 min Pick the tools AI cost simulator

On the job: try it on your own project

Pick one AI feature or agent in your product. Ask Claude Code to check what is logged for each run: prompt, tool calls, token usage, cost, latency and errors. Try to reconstruct one real bad answer from the logs alone. Write a short tracing gap list and the minimum fields every AI call must record, with a note on hiding personal data.

Can you do this now?

0/4. Anything you can't tick yet: reopen that lesson's practice ladder.