agent-ecosystem/test-results/opencode-semantic-model-gauntlet/model-gauntlet-results.json

266 lines
6.4 KiB
JSON

{
"generatedAt": "2026-05-18T13:16:59.867Z",
"runsPerModel": 1,
"qualification": {
"minimumAverageScore": 80,
"minimumSuccessfulRuns": 1,
"minimumConsistencyScore": 85,
"requireNoHardFailures": true
},
"models": [
{
"model": "opencode/big-pickle",
"verdict": "tested-only",
"confidence": "low",
"qualified": false,
"readinessScore": 54,
"averageScore": 35,
"consistencyScore": 100,
"behavioralAverageScore": 35,
"minScore": 35,
"successfulRuns": 0,
"countedRuns": 1,
"hardFailures": 1,
"providerInfraFailures": 0,
"runtimeTransportFailures": 0,
"modelBehaviorFailures": 1,
"harnessFailures": 0,
"p50DurationMs": 133048,
"p95DurationMs": 133048,
"stagePassRates": {
"launchBootstrap": {
"passed": 1,
"total": 1,
"rate": 100
},
"directReply": {
"passed": 1,
"total": 1,
"rate": 100
},
"peerRelayAB": {
"passed": 0,
"total": 1,
"rate": 0
},
"peerRelayBC": {
"passed": 0,
"total": 1,
"rate": 0
},
"concurrentReplies": {
"passed": 0,
"total": 1,
"rate": 0
},
"taskRefs": {
"passed": 1,
"total": 1,
"rate": 100
},
"cleanTranscript": {
"passed": 0,
"total": 1,
"rate": 0
},
"noDuplicateTokens": {
"passed": 0,
"total": 1,
"rate": 0
},
"latencyStable": {
"passed": 0,
"total": 1,
"rate": 0
}
},
"taskRefPassRates": {
"directReply": {
"passed": 1,
"total": 1,
"rate": 100
},
"peerRelayAB": {
"passed": 0,
"total": 0,
"rate": null
},
"peerRelayBC": {
"passed": 0,
"total": 0,
"rate": null
},
"concurrentBob": {
"passed": 0,
"total": 0,
"rate": null
},
"concurrentTom": {
"passed": 0,
"total": 0,
"rate": null
}
},
"protocolViolationTotals": {
"badMessages": 0,
"duplicateOrMissingTokens": 0,
"affectedRuns": 0
},
"stageFailureImpact": [
{
"stage": "concurrentReplies",
"failedRuns": 1,
"weightedLoss": 15,
"passRate": {
"passed": 0,
"total": 1,
"rate": 0
}
},
{
"stage": "peerRelayAB",
"failedRuns": 1,
"weightedLoss": 15,
"passRate": {
"passed": 0,
"total": 1,
"rate": 0
}
},
{
"stage": "peerRelayBC",
"failedRuns": 1,
"weightedLoss": 15,
"passRate": {
"passed": 0,
"total": 1,
"rate": 0
}
},
{
"stage": "cleanTranscript",
"failedRuns": 1,
"weightedLoss": 10,
"passRate": {
"passed": 0,
"total": 1,
"rate": 0
}
},
{
"stage": "latencyStable",
"failedRuns": 1,
"weightedLoss": 5,
"passRate": {
"passed": 0,
"total": 1,
"rate": 0
}
},
{
"stage": "noDuplicateTokens",
"failedRuns": 1,
"weightedLoss": 5,
"passRate": {
"passed": 0,
"total": 1,
"rate": 0
}
},
{
"stage": "directReply",
"failedRuns": 0,
"weightedLoss": 0,
"passRate": {
"passed": 1,
"total": 1,
"rate": 100
}
},
{
"stage": "launchBootstrap",
"failedRuns": 0,
"weightedLoss": 0,
"passRate": {
"passed": 1,
"total": 1,
"rate": 100
}
},
{
"stage": "taskRefs",
"failedRuns": 0,
"weightedLoss": 0,
"passRate": {
"passed": 1,
"total": 1,
"rate": 100
}
}
],
"scoreStability": {
"sampleSize": 1,
"minScore": 35,
"maxScore": 35,
"spread": 0,
"standardDeviation": 0,
"consistencyScore": 100
},
"dominantFailureCategory": "model-behavior",
"recommendationBlockers": [
"overall average 35 < 80",
"behavioral average 35 < 80",
"successful runs 0 < 1",
"hard failures 1",
"model-behavior failures 1",
"highest weighted stage loss concurrentReplies=15"
],
"runs": [
{
"runIndex": 1,
"passed": false,
"score": 35,
"countedForRecommendation": true,
"outcome": "behavioral-fail",
"failureCategory": "model-behavior",
"primaryFailure": "Timed out waiting for OpenCode member bob to become idle. Last durableState: reply_pending",
"durationMs": 133048,
"hardFailure": true,
"stageDurationsMs": {
"setup": 371,
"launchBootstrap": 23735,
"materializeTasks": 36,
"directReply": 17327
},
"stageFailures": {},
"taskRefChecks": {
"directReply": true,
"peerRelayAB": null,
"peerRelayBC": null,
"concurrentBob": null,
"concurrentTom": null
},
"protocolViolations": {
"badMessages": 0,
"duplicateOrMissingTokens": []
},
"stages": {
"launchBootstrap": true,
"directReply": true,
"peerRelayAB": false,
"peerRelayBC": false,
"concurrentReplies": false,
"taskRefs": true,
"cleanTranscript": false,
"noDuplicateTokens": false,
"latencyStable": false
},
"diagnostics": [
"runId=9fb17ac5-fc66-4b01-831e-90a04b1e2304",
"Timed out waiting for OpenCode member bob to become idle. Last durableState: reply_pending"
]
}
]
}
]
}