aggregator committee: deduplicate messages if validator has already been seen for a subnet

[nkryuchkov]

On hold until it's implemented in spec due to broken spec tests
ssvlabs/ssv-spec#603

[nkryuchkov]

@shane-moore @GalRogozinski @MatheusFranco99 #2503 (comment)

[greptile-apps]

Greptile Overview

Greptile Summary

Optimizes sync committee selection proof generation by deduplicating messages when multiple validator sync committee indices map to the same subnet. Since selection proofs depend only on slot and subcommittee index (not the specific validator index), generating multiple signatures for the same subnet is redundant. The change tracks seen subnets and skips duplicate signature generation while maintaining the correct rootToSyncCommitteeIdx mapping needed for later processing.

Confidence Score: 5/5

• This PR is safe to merge once spec tests are updated to match the new behavior
• The logic correctly deduplicates signatures by subnet, which is a valid optimization since selection proofs only depend on slot+subnet. The implementation properly maintains the mapping structure and doesn't introduce any bugs. The PR is on hold waiting for spec changes, not due to code quality issues.
• No files require special attention

Important Files Changed

Filename	Overview
protocol/v2/ssv/runner/aggregator_committee.go	Deduplicates sync committee selection proofs by subnet to avoid redundant signatures for validators mapping to the same subnet

Sequence Diagram

sequenceDiagram
    participant R as AggregatorCommitteeRunner
    participant BN as BeaconNode
    participant S as Signer
    
    Note over R: executeDuty for BNRoleSyncCommitteeContribution
    R->>R: Initialize seenSubnets map
    
    loop For each ValidatorSyncCommitteeIndex
        R->>BN: SyncCommitteeSubnetID(index)
        BN-->>R: subnet
        
        alt subnet already seen
            R->>R: Skip (continue)
        else subnet not seen
            R->>R: Mark subnet as seen
            R->>R: Create SyncAggregatorSelectionData(slot, subnet)
            R->>S: signBeaconObject(data)
            S-->>R: partialSig with SigningRoot
            R->>R: Append to messages
            R->>R: Store rootToSyncCommitteeIdx[SigningRoot] = index
        end
    end
    
    Note over R: Result: One signature per unique subnet instead of per index

Loading

[greptile-apps]

_{1 file reviewed, no comments}

_{Edit Code Review Agent Settings | Greptile}

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
⚠️ Please upload report for BASE (boole-fork@c8c54ee). Learn more about missing BASE report.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

[iurii-ssv]

LGTM, could also mention that this PR pins down the golangci-lint version (in the PR description)

[nkryuchkov]

LGTM, could also mention that this PR pins down the golangci-lint version (in the PR description)

The linter issue actually applies to the base branch as well, so I pulled the latest changes from it

[iurii-ssv]

LGTM, could also mention that this PR pins down the golangci-lint version (in the PR description)

The linter issue actually applies to the base branch as well, so I pulled the latest changes from it

@nkryuchkov I think we have linter pipelines failing for stage branch too since recently, should we create a PR to merge it there too ?

[nkryuchkov]

LGTM, could also mention that this PR pins down the golangci-lint version (in the PR description)

The linter issue actually applies to the base branch as well, so I pulled the latest changes from it

@nkryuchkov I think we have linter pipelines failing for stage branch too since recently, should we create a PR to merge it there too ?

@iurii-ssv #2684