The Starknet team has published a detailed post-mortem following a brief mainnet disruption on Monday, January 5, 2026. The outage, which lasted approximately 18 minutes, was caused by a subtle bug in the network's execution layer, specifically within the blockifier component responsible for processing transactions.
According to the report, the issue stemmed from a mismatch in network state between the blockifier and the separate proving layer, which verifies execution before finalization. In a narrow edge case involving cross-function calls, state writes, and reverts, the blockifier incorrectly retained a state change that should have been discarded. This led to certain transactions being executed incorrectly at the execution layer, even though they should not have passed validation.
"In one specific combination of cross-function calls, variable writes, reverts, and catching them, the blockifier remembered a state-writing that happened within a function that was reverted, causing an incorrect transaction execution," the Starknet team wrote.
Critically, Starknet's safety mechanisms functioned as designed. The proving layer detected the inconsistency and prevented the faulty transactions from being committed to the Ethereum ledger. This forced a block reorganization, rolling back roughly 18 minutes of network activity. The team confirmed that user funds were not compromised and Ethereum finality was never affected.
This incident marks the second major disruption for Starknet in recent months, following a more severe five-hour outage in September 2025 linked to a sequencer bug after the Grinta protocol upgrade. While the latest event was shorter and more contained, both highlight the maturing nature of advanced Layer-2 networks. The Starknet team has committed to expanding testing coverage and audits, particularly around edge-case interactions between execution logic and rollback mechanisms, to reduce the likelihood of similar issues.
