Debug mode

Capture episode workflow logs and Enterprise reflection traces while troubleshooting ingestion.

Debug mode helps you inspect how Zep processes new episodes. When it is enabled, Zep captures detailed workflow logs for episodes processed during the active debug window. Enterprise accounts also capture LLM reflection traces that show how Zep reviewed selected extraction and resolution outputs.

Use debug mode when an episode is pending, a graph update looks unexpected, or you need to understand why ingestion produced a specific result.

Debug logs can include data derived from your episode content and graph processing. Enable debug mode only when you need to troubleshoot, and review logs with the same care as other sensitive operational data.

Enable debug mode

1

Open Project Settings

In the Zep dashboard, open the project you want to debug, then go to Project Settings.

2

Find Debug Options

In Debug Options, click Enable Debug Logging.

3

Process new episodes

Add or reprocess the episodes you want to inspect while debug mode is active. Debug mode does not backfill logs for episodes processed before it was enabled.

4

Open episode logs

Go to a user or graph episode list. Open the episode action menu and select Debug Logs.

Debug mode is scoped to the project where it is enabled. It is time-limited and automatically turns off:

  • Standard accounts: 60 minutes.
  • Enterprise accounts: 24 hours.

Enabling debug mode again resets the active window.

Workflow logs

The Workflow tab in the episode logs dialog shows the ingestion workflow for that episode.

Workflow logs include:

  • Episode status, start time, completion time, and top-level error details.
  • Timestamped log entries with level and message.
  • Step-by-step ingestion progress through graph construction.
  • Errors and stack traces when a workflow step fails.
  • Processing details that can help explain why a node, edge, or fact was created, changed, skipped, or failed.

Logs are available only for episodes processed while debug mode was active. Workflow logs are retained for a limited time: 12 hours by default and 7 days for Enterprise accounts.

Per-episode debug logs are not currently supported for batch ingestion operations.

Reflection traces

Reflection tracing is available only for Enterprise accounts. Turning on debug mode for an Enterprise project also enables reflection tracing for eligible ingestion steps during the active debug window.

Reflection traces show how a reviewer model inspected selected LLM outputs during graph construction. The reviewer can apply corrections before downstream steps run.

The Reflection tab can include:

  • Stage and prompt name, such as extraction or resolution.
  • Validation status, including whether the reviewed output was applied or failed validation.
  • Model and model size used for the review.
  • Attempt count and error details when retries were needed.
  • Explanations for the reviewer’s decisions.
  • Corrections, warnings, and metadata.
  • Original output and reviewed output for comparison.

Reflection traces are most useful when an extracted entity, edge, or fact looks wrong. They let you compare the primary model output with the reviewed output and see why the reviewer accepted, corrected, or rejected part of the result.

Troubleshooting

If an episode has no logs, check that debug mode was enabled before the episode was processed and that the log retention window has not expired.

If the Reflection tab is empty, one of the following is usually true:

  • The account is not Enterprise.
  • The episode was processed while debug mode was off.
  • The episode did not run an ingestion step that emits reflection traces.
  • Reflection tracing failed or exhausted retries. In that case, ingestion continues with the primary output.

Debug mode does not change retrieval behavior for existing graph data. It affects logging and, for Enterprise accounts, enables reflection review during new ingestion workflows.