Enable the siUSD pause-to-CRITICAL rule #441

Epic: #437

Outcome

a valid included siUSD pause invokes the existing infiniFi CRITICAL automation.

Scope

• Map the exact chain/emitter/event to protocol=infinifi, severity CRITICAL, deterministically.
• Act at first inclusion. Handle event removals, duplicate sources, and re-inclusion through
  incident provenance and action IDs, without automatically undoing completed protective actions.
• Agree startup, historical replay, and recovered-during-outage policies. Reconcile paused()
  at startup; unpause records recovery without automatically restoring exposure.
• Verify the configured siUSD/PT-siUSD vault scope across Ethereum and Katana.
• Validate the whole flow using a local fork/non-executing environment, then live observation
  with actions disabled, before enabling this one rule. Update protocol docs and monitoring metadata.

Acceptance criteria

the rule passes failure/reorg/recovery scenarios, produces one accepted emergency
action per incident, and measured latency meets the agreed target. Report acceptance, first
transaction submission, and completion separately; ingestion alone cannot guarantee exit speed.

Dependencies

Blocked by:

• Move timelock alert delivery to the continuous worker #439
• Index siUSD pause and unpause events yearn-envio#69
• https://github.com/tapired/liquidity-monitoring/issues/452
• Deliver critical requests independently of Telegram #440

Related work

• Existing infiniFi monitor documentation: https://github.com/yearn/monitoring/blob/main/protocols/infinifi/README.md
• Execution scope: https://github.com/tapired/liquidity-monitoring/blob/main/src/hooks/emergency_config.json

Include relevant behavior/failure tests and operational documentation in the implementation. Update monitoring metadata when monitor behavior changes.