Skip to main content
Once a workflow is live, you’ll want to know it’s actually doing its job — sending the right messages, to the right people, without errors. This guide shows you how to read a workflow’s Logs and its analytics so you can trust it and spot problems early.

​
Before you begin

  • The workflow should be turned On and have had a chance to run (its trigger has fired at least once).
  • You’ll need permission to manage your AI agent (Admin or Manager).
The Logs answer “what happened on this run?” The analytics answer “how is this workflow performing overall?” Use both.
1

Open the Logs tab

  • On the workflow’s page, click the Logs tab.
  • You’ll see a list of individual runs — one per time the workflow fired — newest first.
2

Read each run's status

Every run carries a status that tells you exactly what happened:
  • Running — the run is in progress right now.
  • Waiting — the run is paused mid-way (for example, waiting until the day before an appointment).
  • Completed — the workflow finished all its steps successfully.
  • Contact replied — the run was cancelled because the customer answered, so no further automated message was needed.
  • Policy blocked — a message couldn’t be sent because the channel’s rules didn’t allow it at that moment.
  • Failed — something went wrong during the run.
You can now see, run by run, what your workflow is doing.
3

Open a run for details

  • Click a run to expand its step-by-step log.
  • You’ll see each action it took and when — useful for understanding a Policy blocked or Failed result.
  • When a run is blocked, the log spells out the reason in plain language (for example, Contact opted out, No channels selected, or Trigger fired on a different channel) so you know what to change.
Each messaging platform sets its own rules for when you can message a customer. LINE and Telegram stay open, while Messenger and Instagram allow free-form replies for about 24 hours after the customer’s last message (extended when a teammate is handling the chat), and WhatsApp allows about 24 hours. When an action falls outside a platform’s window, Cura records the run as Policy blocked rather than risking your account.
4

Check the analytics

The workflow’s analytics summarize performance over time, so you don’t have to count runs by hand.
  • Triggered — how many times the workflow started.
  • Completed — how many runs finished successfully.
  • Success — the share of runs that completed, shown as a percentage.
  • Policy blocked — how many runs were stopped because a channel wouldn’t allow the message.
  • Switch the range between 7d, 30d, and 90d to see recent trends. It opens on 30d by default.
A healthy reminder workflow shows a high Success rate. A rising Policy blocked count usually means you’re trying to message customers outside what their channel allows — adjust timing or scope.

​
How to know it’s working

  • After a real trigger (e.g. a new booking on a scoped channel), a fresh run appears in Logs and ends in Completed.
  • The Triggered and Completed counts in analytics climb together over your chosen range.

​
Troubleshooting

  • No runs at all. The trigger may not have fired, or no channel is scoped. Confirm the workflow’s On / Off toggle is set to On and check its Apply to tab.
  • Many Policy blocked runs. The channel won’t allow messages at that point. Review your timing and which channels are scoped.
  • A Failed run. Open it to read the step log, then adjust the configuration and test again.