Public leaderboard data for completed gas optimization reports.

Gas Leaderboard

The public Gas Leaderboard lives at:

/gas-leaderboard

It ranks completed Archon gas reports from the database. It does not run new analysis and it does not invent placeholder rows. If production has no matching reports, the page shows an empty state and links users to the Gas Optimizer.

The JSON data endpoint is:

GET /api/gas/leaderboard

Both surfaces are public and do not require an app session.

What appears on the page

Each row links to its public gas report at /app/gas/:gasReportId and shows:

  • rank,
  • contract name,
  • source label,
  • source hash,
  • report date,
  • traffic assumptions,
  • gas efficiency score,
  • annualized savings,
  • L2 gas saved per call,
  • L1 / data-availability wei saved per call,
  • proof link when the gas report has an anchorTxHash.

Source labels are explicit:

  • Sample · Archon-labeled means Archon's own sample contract, not a third-party deployment.
  • Verified address means the report came from address mode and stores the source reference.
  • User-submitted scan means pasted or uploaded source.

Filters

The page and API support the same core filters.

ParameterValuesDefaultNotes
metricscore, savings, l2, recentscoreSelects sort order.
sourceKindall, sample, paste, addressallFilters by stored report source kind.
qtext up to 120 charsemptySearches contract name, source reference, or source hash.
limit1-10050API-only cap for returned rows. The page shows up to 50.

Invalid API query values return 400 with validation issues.

Response shape

GET /api/gas/leaderboard returns:

  • schema: currently archon.gas.leaderboard.v2,
  • generatedAt: response timestamp,
  • filters: normalized query filters,
  • assumption: comparison limits,
  • rows: leaderboard entries.

Each row includes:

  • gasReportId, contractName, sourceKind, sourceRef, sourceHash,
  • reportHash and anchorTxHash when proof data exists,
  • totals and assumptions,
  • optimizationCount and measuredOptimizationCount,
  • annualSavingsUsd, l2GasSavedPerCall, l1DaWeiSavedPerCall,
  • gasEfficiencyScore,
  • createdAt.

Ranking rules

metric=score orders by a computed gas efficiency score, then annual savings, then recency. The score combines:

  • stored annual USD savings,
  • L2 gas saved per call,
  • optimization count, capped to avoid rewarding noisy reports too much,
  • small bonus for anchored proof reports.

metric=savings prioritizes annual USD savings. metric=l2 prioritizes L2 gas saved per call. metric=recent shows newest completed reports first.

Data rule

Rows must come from stored, completed reports (status='done'). Annual savings use each report's stored callsPerYear and MNT/USD assumptions, so compare rows only after normalizing traffic assumptions.

If a value is estimated rather than measured, downstream consumers should label it as estimated instead of blending it with receipt-backed measurements. The leaderboard exposes measuredOptimizationCount so integrations can distinguish measured-heavy reports from estimate-heavy reports.