Failed executions are highlighted in red in the table. Use the Errors filter button above the table to show only failed executions. You can also filter by time range using the time selector, or search by pipeline name.

The detail panel shows execution metadata including the pipeline path, Snaplex node name, execution duration, and the error message with resolution hint. The Snap statistics tab is selected by default, showing per-Snap duration, CPU, memory, and document counts.

The panel header shows a warning icon and the name of the pipeline being analyzed. The result contains two sections:

• Root cause: A concise headline describing the probable cause of the failure, followed by a detailed explanation. Technical terms, class names, and code fragments are formatted inline for readability. Click the section header to collapse or expand it.
• Suggested fixes: A prioritized list of recommended actions to resolve the issue. Each fix card shows:
  • A Recommended badge on the primary fix.
  • A title and description of the fix.
  • Where applicable, a code diff showing the change to make, with lines to remove and lines to add clearly marked.
  • An Open in Designer button. The button on the recommended fix is filled; the buttons on other fixes are outlined.

A disclaimer at the bottom of the panel reads: AI may make mistakes. Verify suggested fixes before applying.

Designer opens in a new browser tab and displays the Pipeline Execution Statistics dialog for the failed execution. The dialog shows the failing Snap with a red error indicator, the top-level error message, and the reason for the failure. Use the Snap Statistics, Pipeline Parameters, and Extra Details tabs to inspect the execution in detail. Close the dialog to access the pipeline canvas.

The SnapGPT Assistant panel remains open on the Monitor page with the analysis visible, so you can refer to the suggested fix while working in Designer.