@lythos/skill-arena 0.9.8 → 0.9.9

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (2) hide show
  1. package/package.json +1 -1
  2. package/src/runner.ts +8 -0
package/package.json CHANGED
@@ -1,6 +1,6 @@
1
1
  {
2
2
  "name": "@lythos/skill-arena",
3
- "version": "0.9.8",
3
+ "version": "0.9.9",
4
4
  "description": "Skill Arena — benchmark skill effectiveness with controlled-variable comparison",
5
5
  "keywords": [
6
6
  "ai-agent",
package/src/runner.ts CHANGED
@@ -119,6 +119,14 @@ export async function runArenaFromToml(opts: {
119
119
  criteria: [],
120
120
  }) as JudgeVerdict
121
121
 
122
+ // Persist per-cell verdict + agent output for auditability
123
+ writeFileSync(join(cellDir, 'judge-verdict.json'), JSON.stringify({
124
+ ...v,
125
+ agent_stdout: result.agentResult.stdout.slice(0, 5000),
126
+ agent_stderr: result.agentResult.stderr.slice(0, 1000),
127
+ duration_ms: result.agentResult.durationMs,
128
+ }, null, 2) + '\n')
129
+
122
130
  if (!verdictsBySide.has(cell.side)) verdictsBySide.set(cell.side, [])
123
131
  verdictsBySide.get(cell.side)!.push(v)
124
132
  } catch (e) {