Reviewing files that changed from the base of the PR and between d610ec1 and a023a80.

• lib/urc (1 hunks)
• src/libs/SlashingLib.sol (1 hunks)
• src/slasher/LinglongSlasher.sol (1 hunks)
• test/EigenLayerMiddleware.t.sol (1 hunks)
• test/SymbioticMiddleware.t.sol (1 hunks)

• GitHub Check: Foundry Project

lib/urc (1)

1-1:

Details

❓ Verification inconclusive

Submodule update looks correct; verify for breaking changes.

The lib/urc subproject pointer has been bumped to commit 5f18225c4d027b49084605129b7666a37ea13412. No public interface changes are expected, but please confirm that:

• The new commit aligns with the intended timestamp-based logic updates.
• All dependent tests in lib/urc still pass.
• There are no unintended API changes used by the slashing library.

---

🏁 Script executed:

#!/bin/bash
# Fetch and diff the submodule changes
git submodule update --init lib/urc
cd lib/urc
git fetch
git diff e0c5c8a8212a04e8844df0268c7de17b9e6b8896 5f18225c4d027b49084605129b7666a37ea13412

# Run the submodule's test suite (if available)
# e.g., cargo test || npm test

Length of output: 70062

---

Confirm lib/urc submodule API & logic updates

The bump to commit 5f18225… introduces the following core changes in lib/urc:

• Switched all on-chain timing from block.number → block.timestamp in Registry.sol.
• Updated IRegistry & ISlasher interfaces:
  • Timestamp-based fields now use uint48/uint24 widths.
  • Removed the old slashFromOptIn method.
  • Added a committer parameter to the slash signature.
• Adjusted example contracts (InclusionPreconfSlasher.sol, StateLockSlasher.sol) and scripts to match the new API.
• Added config/registry.json, expanded scripts folder, and updated Foundry settings.

Please verify that:

• All downstream code (clients, slashing library, examples) has been updated to call the new slash(..., committer, …) signature and no longer relies on slashFromOptIn.
• There are no unintended API changes breaking upstream usage.
• All lib/urc tests still pass (e.g. via forge test or npm test).
• Integration/slashing tests in the main repo remain green.

src/libs/SlashingLib.sol (1)

308-310:

Details

❓ Verification inconclusive

Possible unit-mismatch in fraud-proof window check

registeredAt is still stored as a block number elsewhere in the library (unregisteredAt is compared with block.number a few lines above).
By switching this clause to block.timestamp you implicitly assume that registeredAt is now a Unix-timestamp. If that field still contains a block number, the comparison mixes blocks with seconds and the fraud-proof window will be orders of magnitude longer than intended.

Action items
• Double-check how OperatorData.registeredAt is written in Registry after this change.
• If it is still a block number, convert it to seconds (or vice-versa) before the addition.
• Consider migrating every *_At field to a single time base to avoid future bugs (the adjacent unregisteredAt clause still uses block.number!).

Example patch if timestamps are desired everywhere:

// Treat all *_At fields as timestamps (seconds)
-        if (unregisteredAt < block.number) {
+        if (unregisteredAt < block.timestamp) {
             revert OperatorUnregistered();
         }

-        if (registeredAt + registry.getConfig().fraudProofWindow > block.timestamp) {
+        if (uint256(registeredAt) + registry.getConfig().fraudProofWindow > block.timestamp) {
             revert OperatorFraudProofPeriodNotOver();
         }

---

🏁 Script executed:

#!/bin/bash
# Inspect every assignment to `registeredAt` to confirm whether
# a timestamp or block number is stored.
rg --no-heading -n '\.registeredAt\s*='

Length of output: 146

---

🏁 Script executed:

#!/bin/bash
# Retrieve the OperatorData struct definition to check types of registeredAt and unregisteredAt
rg --no-heading -n 'struct OperatorData' -A10

# Search for assignments to registeredAt across the repository to see if it's set to block.number or block.timestamp
rg --no-heading -n 'registeredAt\s*=' -g '*.sol' -A3

Length of output: 1889

---

🏁 Script executed:

#!/bin/bash
# Locate the OperatorData struct definition and its field types
rg --no-heading -n 'struct OperatorData' -g 'src/**/*.sol' -A15

# Find the implementation of getOperatorData to see where OperatorData lives
rg --no-heading -n 'function getOperatorData' -g 'src/**/*.sol' -A15

# Search for where operatorData.registeredAt is assigned in registry code
rg --no-heading -n 'operatorData\.registeredAt\s*=' -g 'src/**/*.sol' -A5

Length of output: 212

---

🏁 Script executed:

#!/bin/bash
# Locate all references to OperatorData to find its struct definition and type of registeredAt
rg --no-heading -n 'OperatorData' -g 'src/**/*.sol'

# Fallback grep in case rg misses it
grep -R -n 'struct OperatorData' src

Length of output: 1114

---

Verify OperatorData.registeredAt time unit before merging

I wasn’t able to locate the struct OperatorData definition in this repo (it lives in the external IRegistry interface), so please double-check whether its registeredAt field is a block number or a Unix-timestamp. Mixing blocks with seconds here can drastically extend the fraud-proof window if you assume the wrong unit.

Action items
• Confirm the exact type/unit of IRegistry.OperatorData.registeredAt in the registry contract.
• If it’s still stored as a block number, convert it to seconds (or switch the comparison to block.number) before adding fraudProofWindow.
• Consider standardizing all *_At fields (e.g. registeredAt, unregisteredAt, slashedAt) on the same time base to prevent similar bugs in the future.

src/slasher/LinglongSlasher.sol (1)

215-219: Great improvement to function signature readability

Changing from unnamed to named parameters in the slash function significantly improves code readability and maintainability. This change aligns with best practices for interface implementations, making it clearer to developers what each parameter represents while maintaining the same functionality.

Reviewing files that changed from the base of the PR and between b7c9e0e and ca232ad.

• script/Deployments.s.sol (1 hunks)
• src/slasher/LinglongSlasher.sol (2 hunks)
• test/EigenLayerMiddleware.t.sol (2 hunks)
• test/utils/EigenlayerDeployer.sol (1 hunks)
• test/utils/MockChallenger.sol (1 hunks)

• script/Deployments.s.sol

• test/EigenLayerMiddleware.t.sol

test/utils/EigenlayerDeployer.sol (1)

111-111: Making MIN_WITHDRAWAL_DELAY public improves test accessibility

Changing the visibility modifier from internal to public is appropriate here. This allows external contracts and test code to directly access this constant value, which facilitates more precise control and readability in test scenarios involving withdrawal timing logic.