Battle Arena

Compare two skills under identical conditions with multi-factor scoring.

The Battle Arena compares two skills using the same input and runtime conditions.

It gives users a structured way to evaluate alternatives.

How a battle works

1. Select two skills

2. Define the shared protocol and action context

3. Execute both skills simultaneously

4. Score both outputs across six dimensions

5. Determine the winner by total score

6. Save the result for history and analysis

Scoring dimensions

Each contestant is scored across:

• Speed

• Depth

• Clarity

• Protocol coverage

• Docs quality

• Execution readiness

Example battle request

{
  "contestant1": { "id": 1, "name": "Jupiter Swap", "slug": "jupiter-swap" },
  "contestant2": { "id": 2, "name": "Orca Swap", "slug": "orca-swap" },
  "protocol": "jupiter",
  "action": "swap",
  "battleType": "standard"
}

Example battle response

{
  "winner": { "id": 1, "name": "Jupiter Swap", "slug": "jupiter-swap" },
  "contestant1Score": {
    "speed": 9.1,
    "depth": 8.7,
    "clarity": 9.0,
    "protocolCoverage": 8.5,
    "docsQuality": 9.2,
    "executionReadiness": 8.8
  },
  "contestant2Score": { "...": "..." },
  "winnerName": "Jupiter Swap",
  "latencyMs": 421
}

Battle history

Recent results can be queried from the history endpoint.

History entries can show expanded score breakdowns for both contestants.

Why Arena matters

The Arena gives teams a repeatable benchmarking surface.

It reduces guesswork when choosing between skills or protocols.