Trace, evaluate & monitor runs

LESSONLesson 5 · ~25 min

🎯Goal. See exactly what your agent did on each run — every step, its inputs, timing and token cost — and evaluate quality over time.

▶ Try this prompt

Open the trace for the last run and check which agent handled each step, how long it took, and how many tokens it used.

Flowise includes tracing/analytics and Evaluations & Metrics (the latter is on every plan, including Free) for monitoring runs.

1Run a multi-step Agentflow, then open its execution trace — a step-by-step view (Start → workers → final answer) showing each agent's input and output.
2Read the per-step timing and token counts to find slow or expensive steps; this is how you debug a flow that gives a wrong or costly answer.
3Use Evaluations & Metrics to score runs against expected answers so you can tell whether a change to the flow actually improved it — not just feels better.

A Flowise execution trace listing Supervisor and worker steps down the left with a Generate Final Answer step selected, showing timing and token counts — An **execution trace** in Flowise: each step of a multi-agent run (Supervisor → workers → Generate Final Answer) with its inputs, duration and token count — observability for debugging and cost. Source: https://flowiseai.com

✓You'll see. A trace listing every step of a run with its inputs, duration and token usage — a clear record of what the agent actually did.

💳Cost. Tracing, analytics and Evaluations & Metrics are included on the Free plan; running the evaluations themselves spends normal model tokens.

💡Takeaway. You can't improve what you can't see — traces and evaluations make an agent's behaviour and cost measurable.