Certificate anchors auto-recover from transient failures

Failed anchors automatically retry when the blockchain service recovers, eliminating manual intervention.

Bug Fixes

Previously, when a certificate anchor failed to submit due to temporary network issues or service unavailability, the system had no way to recover—the anchor would remain stuck in a failed state indefinitely. Now the automatic reconciliation process retries failed anchors that don't have a server-side record, allowing them to successfully complete on recovery. This applies to both certificate issuances and transfers.