Automation Retry Policy

Defines Krystal’s retry mechanism for automation systems.

Overview

Automation systems on Krystal (such as Auto-Rebalance, Auto-Exit, Auto-Compound, and Auto-Harvest) include a built-in retry mechanism to improve reliability and ensure that scheduled actions are executed successfully, even under temporary network or on-chain disruptions.

Retry Rules

Parameter	Value	Description
Max Retry Attempts	12	Each automation order may be retried up to 12 times if execution fails due to transient errors.
Retry Interval	5 minutes	Between each retry, the system waits 5 minutes before re-attempting execution.
Total Retry Duration	~60 minutes	In total, the system continues retrying for approximately 1 hour before the task is marked as failed.

Design Rationale

This retry mechanism improves automation reliability under fluctuating network conditions while maintaining a bounded execution window. The 12×5-minute retry policy ensures:

• Timely completion under normal conditions

• Avoidance of indefinite looping or spam retries

• Predictable failure window (~1 hour)

After Retry Window

If an automation still fails after the full retry period (~60 minutes), it is marked as failed. To restore normal operation, users should:

• Delete and re-create the automation, or

• Edit and update the existing automation to trigger a new execution cycle.

This ensures the automation restarts with the latest market and pool data.