Oracle, Liquidation, and MEV Risk

Oracle integrations are not safe just because the provider is reputable. Teams still need feed selection, stale-value handling, decimals, sequencer status, fallbacks, liquidation stress tests, and circuit breakers.

Review questions

• What happens when the feed is stale, delayed, wrong, missing, or unavailable?
• Are liquidations economically viable during volatile markets and gas spikes?
• Can transaction ordering, sandwiching, backrunning, or stale updates move value?

Review workflow

1. List feed sources, decimals, heartbeat, deviation, sequencer status, and fallback behavior.
2. Stress test solvency, liquidation, and keeper incentives under volatility.
3. Simulate MEV-sensitive paths before launch.
4. Add circuit breakers that operators can actually execute and restart safely.

Common risks

Risk	What to verify
Stale values	Heartbeat, age checks, fallback, and sequencer status.
Decimal mismatch	Unit tests for every feed and asset path.
Keeper failure	Incentive, gas, liquidity, and backlog stress tests.
Unsafe restart	Pause, resume, recovery, and accounting reconciliation.

Linked checklists

• Oracle, liquidation, and MEV readiness
• Launch readiness
• Post-launch operations

FAQ

Where does MEV enter the review? Review swaps, liquidations, oracle updates, auctions, withdrawals, perps funding, and privileged operations that depend on ordering.

What should a launch gate require? Feed assumptions, stale-data tests, stress simulations, keeper failure handling, and operable circuit breakers.

Starting resources

• Chainlink feed selection docs
• Tenderly docs
• BlockSec Phalcon simulator