Incident, keeper failure

Trigger: vigil.keeper.miss.count > 0, or fewer than 2 of 3 keepers reporting healthy.

Response

1. Check which keeper is down. praetor keepers list shows last action block per keeper.
2. Investigate root cause:

• VPS health: SSH in, check process logs, disk space, memory.
• RPC: is the RPC URL responsive?
• Wallet: does the keeper wallet have enough native ETH for gas?
• Stake: is the keeper still meeting keeper_min_stake_wei?

1. Bring the keeper back:

• Restart the keeper binary.
• Watch for the next KeeperStaked or successful LiquidationExecuted event.

1. If the keeper is unrecoverable:

• praetor keepers slash --keeper <addr> --reason "unrecoverable_failure"
• Bring up a replacement keeper on a new VPS.
• Update Cohort partners that one keeper rotated.

1. If fewer than 2 keepers are healthy at any time:

• Praetor multisig invokes Plinth.pause("keeper redundancy lost") until at least 2 are back.
• Withdrawals still work; new positions are paused.

1. Communicate via Lantern dashboard banner and Mirror post.
2. Post-mortem within 7 days.