Logs

Logs show what happened in each workflow run so you can debug faster and operate with confidence.

How to Access Logs

From the workflow canvas, click the Logs tab in the left panel. You’ll see a list of recent runs with status, duration and timestamp.

What Logs Include

Each run entry shows:

• Timestamp — when the run started
• Status — completed, failed, or running
• Duration — total time from trigger to final output
• Cost signals — token usage and estimated cost per LLM block
• Input and output context — the payload sent in and the data returned
• Error messages — if the run failed, which block failed and why

Click any run to expand the block-by-block trace. This shows the output of each block in sequence, so you can see exactly where things went wrong.

Filter Strategy

Use filters to find issues quickly:

• Date range for incident windows
• Status = failed for active debugging
• Status = completed for performance review

Failure Triage Pattern

1. Find the first failing run in your time window
2. Open the failing block details
3. Identify the failure type:
   • Input error — the data coming into the block was missing, malformed or the wrong type. Check the input context on the run and trace back to your trigger or an earlier block’s output.
   • Logic error — the block received good data but produced wrong output. Usually a template bug or misconfigured step. Reproduce in Console with the same input.
   • External dependency error — a third-party API or integration returned an error or timed out. Check the error message for status codes and retry after the dependency recovers.
4. Patch the workflow and test in Console
5. Re-run and confirm recovery

Operations Tips

• Watch duration trends for performance regressions — a block that normally takes 200ms jumping to 2s is a signal worth investigating
• Track recurring failure signatures by block type to spot systemic issues
• Pair Logs with History to map failures to specific revisions

Next Steps

• Console: Reproduce failures with controlled input
• History: Identify the change that introduced failure
• Environments: Roll back or redeploy safely

Built with ❤️ by Scout OS