Debugging Executions

puq.ai logs every workflow run with full visibility into each step of the execution.
This allows you to understand what happened, identify errors, and fix issues quickly — even in complex automations.

Debugging tools help you verify correctness as you build and maintain your workflows.

Where to View Executions

You can access execution logs from two places:

1. Global Executions Dashboard

Navigate to:
Dashboard → Executions

This view shows:

  • All executions across every workflow
  • Their statuses (Success, Failed, Running…)
  • Trigger source
  • Execution time
  • Ability to open detailed logs

Use this to monitor system-wide automation activity.

2. Workflow Editor → Debug Panel

When editing a workflow, open the Debug tab.
This panel shows:

  • Executions only for the current workflow
  • Recent runs with timestamps
  • Quick links to inspection mode

Perfect for debugging while building.

What You Can Inspect

Each execution provides a rich set of debugging tools:

🔍 Step-by-Step View

See the workflow exactly as it ran — node by node.

📥 Input / 📤 Output Inspection

For every node, you can view:

  • Raw inputs received
  • Outputs generated
  • Transformed data
  • Branching decisions in Routers

This makes it easy to trace how data moves through your automation.

❗ Error Messages & Stack Traces

Failed executions show:

  • Error message
  • Stack trace (if available)
  • The node where the failure occurred
  • Suggestions for recovery

Great for quickly locating misconfigured nodes or invalid data.

🔁 Retry Failed Executions

You can retry any failed run instantly.
Retries use the same trigger data, making debugging deterministic and repeatable.

Live Testing Mode

Use Run Step to verify a single step’s behavior while building a workflow, without running everything after it.

Run Step allows you to:

  • Trigger a test run of the workflow up to the selected step
  • See that step’s real input and output as formatted JSON
  • Validate node outputs step by step
  • Catch misconfigurations before publishing

This dramatically reduces debugging time.