RESULTS

Generated by pnpm bench from bench/results/*.json. Do not edit by hand (Hard Rule 12).

core

Measured 2026-10-13 with pnpm bench.

Metric Value
observations 50000
claims 5000
runs 5
bucketMinutes 15
lookupMedianMs 446.9
lookupMinMs 379.3
lookupMaxMs 488.5
lookupCpuMedianMs 467.3
indexBuildMedianMs 166.6
buildPlusLookupMedianMs 619.6
candidatesReturned 383354
gateMs 2000
gateMetric lookupCpuMedianMs
gatePassed true
gradingMedianMs 1242.2
gradingMinMs 1176.9
gradingMaxMs 1498
gradingCpuMedianMs 1446.9
gradingReceipts 5803
gradingVerified 1883
gradingNotFound 910
gradingUnasked 802
gradingGateMs 2000
gradingGateMetric gradingCpuMedianMs
gradingGatePassed true
commit 23ea9c6b689b37f3ab74dd64ea2c70d16177e754
commitDirty false
node v22.14.0

gcal

Measured 2026-10-13 with pnpm bench.

Metric Value
gcalSelectorsVersion 1.5.0
gcalFixtures 18
gcalSyntheticFixtures 8
gcalEventsLabelled 67
gcalEventsFound 67
gcalEventsSpurious 0
gcalEventFieldsChecked 335
gcalEventFieldsCorrect 335
gcalStartAccuracy 1
gcalAllDayAccuracy 1
gcalSubjectAccuracy 1
gcalGuestsAccuracy 1
gcalCalendarAccuracy 1
gcalHealthCorrect 18/18
gcalCoverageWindowsCorrect 17/17
gcalAllCalendarsCorrect 17/17
gcalStepDiffsExact 12/12
gcalStepRowsCorrect 6/6
gcalStepRowsSpurious 0
gcalStepRowsUnmarked 0
gcalAccuracy 1
gcalAccuracyThreshold 0.95
gcalAccuracyGatePassed true
commit 23ea9c6b689b37f3ab74dd64ea2c70d16177e754
commitDirty false
node v22.14.0

gmail-sent

Measured 2026-10-13 with pnpm bench.

Metric Value
gmailSentSelectorsVersion 1.4.0
gmailFixtureFolders 2026-10-03,2026-10-06,2026-10-07
gmailFixtures 41
gmailSentListFixtures 11
gmailSentRealFixtures 4
gmailSentLabelledObservations 108
gmailSentProducedObservations 108
gmailSentFieldsCompared 324
gmailSentFieldsCorrect 324
gmailSentRecipientAccuracy 1
gmailSentSubjectHintAccuracy 1
gmailSentOccurredAtAccuracy 1
gmailSentAccuracy 1
gmailSentAccuracyGate 0.95
gmailSentAccuracyGatePassed true
gmailSentHealthGatePassed true
gmailSentNegativesGatePassed true
gmailSentCoverageGatePassed true
gmailSentDedupGatePassed true
gmailSentSubjectHintMaxCodePoints 80
gmailSentSubjectHintGatePassed true
gmailSentSeedsGatePassed true
commit 23ea9c6b689b37f3ab74dd64ea2c70d16177e754
commitDirty false
node v22.14.0

intel

Measured 2026-10-13 with pnpm bench.

Metric Value
intelCases 24
intelLabelled 14
intelAsked 17
intelRulesVerifiedTrue 5
intelRulesVerifiedFalse 0
intelRulesPrecision 1
intelRulesRecall 0.3571
intelBuiltinMeasured false
intelByoAnthropicMeasured false
intelByoOpenaiMeasured false
intelModelHelps false
intelCorpusInvented true
commit 23ea9c6b689b37f3ab74dd64ea2c70d16177e754
commitDirty false
node v22.14.0

paste

Measured 2026-10-13 with pnpm bench.

Metric Value
pasteLogs 4
pasteFields 111
pasteFieldAccuracy 1
pasteAccuracyThreshold 0.95
pasteCorpusInvented true
pasteGatePassed true
commit 23ea9c6b689b37f3ab74dd64ea2c70d16177e754
commitDirty false
node v22.14.0

patterns

Measured 2026-10-13 with pnpm bench.

Metric Value
museSentences 68
museTypeAccuracy 0.9559
museWrongActions 0
chatgptSentences 67
chatgptTypeAccuracy 0.9552
chatgptWrongActions 0
dotsSentences 67
dotsTypeAccuracy 0.9552
dotsWrongActions 0
claudeSentences 67
claudeTypeAccuracy 0.9552
claudeWrongActions 0
patternsAccuracyThreshold 0.9
patternsCorpusInvented true
patternsGatePassed true
commit 23ea9c6b689b37f3ab74dd64ea2c70d16177e754
commitDirty false
node v22.14.0

perf

Measured 2026-10-13 with pnpm bench.

Metric Value
perfGmailRows 10000
perfGmailPassMedianMs 776.2
perfGmailPassMinMs 763.4
perfGmailPassMaxMs 815.6
perfGmailPassGateMs 2000
perfGmailPassGatePassed true
perfGmailStoreMs 90345
perfGmailRowsAfterTwoReads 10000
perfGmailSentWindows 1
perfMonthClaims 5000
perfMonthObservations 50000
perfMonthStorageBytes 40208677
perfMonthStorageGateBytes 50000000
perfMonthStorageGatePassed true
perfMonthPopupFirstPaintMedianMs 60
perfMonthPopupGateMs 100
perfMonthPopupGatePassed true
perfExportJsonMs 3530
perfExportJsonBytes 40143898
perfExportJsonPeakHeapBytes 59278905
perfExportCsvMs 1251
perfExportCsvBytes 1206053
perfExportCsvPeakHeapBytes 32115739
perfWorkerIdleHeapBytes 3459608
perfMonthSealJobMs 20985
perfMonthSealAndRebuildMs 32373
perfMonthSealedStoragePeakBytes 67749026
perfMonthSealedStorageBytes 59004244
perfMonthSealedPopupFirstPaintMedianMs 76
perfMonthSealedPopupGatePassed true
perfMonthSealedExportJsonMs 7048
commit 23ea9c6b689b37f3ab74dd64ea2c70d16177e754
commitDirty false
node v22.14.0

Measured 2026-10-13 with pnpm bench.

Metric Value
popupReceipts 5000
popupRuns 5
popupRows 50
popupFirstPaintMedianMs 60
popupFirstPaintMinMs 52
popupFirstPaintMaxMs 64
popupDayPaintMedianMs 112
popupDayPaintMinMs 100
popupDayPaintMaxMs 112
popupCpuMedianMs 84.9
popupCpuMinMs 77.5
popupCpuMaxMs 101
popupChromeMajor 153
popupTrace bench/traces/popup-paint.zip
popupPaintGateMs 100
popupPaintGateMetric popupFirstPaintMedianMs
popupPaintGatePassed true
commit 23ea9c6b689b37f3ab74dd64ea2c70d16177e754
commitDirty false
node v22.14.0

receipts

Measured 2026-10-13 with pnpm bench.

Metric Value
gmailFixtures 41
receiptRealFixtures 6
stripeFixtures 3
stripeExact 3
stripeFieldAccuracy 1
amazonFixtures 1
amazonExact 1
amazonFieldAccuracy 1
marketplaceFixtures 1
marketplaceExact 1
marketplaceFieldAccuracy 1
opentableFixtures 1
opentableExact 1
opentableFieldAccuracy 1
instacartFixtures 2
instacartExact 2
instacartFieldAccuracy 1
uberFixtures 2
uberExact 2
uberFieldAccuracy 1
shopifyFixtures 1
shopifyExact 1
shopifyFieldAccuracy 1
genericFixtures 4
genericExact 4
genericFieldAccuracy 1
receiptFieldAccuracy 1
negatives 7
negativesUnparsed 7
listViews 15
listViewsNeutral 15
coveragePages 41
coveragePagesExact 41
coverageWindows 11
healthLabelled 26
healthMatched 26
receiptParsersGatePassed true
commit 23ea9c6b689b37f3ab74dd64ea2c70d16177e754
commitDirty false
node v22.14.0

redteam

Measured 2026-10-13 with pnpm bench.

Metric Value
redteamStrategies 13
redteamCases 131
redteamWorking 6
redteam01ClaimInflationCases 5
redteam01ClaimInflationWorks 0
redteam01ClaimInflationRate 0
redteam01ClaimInflationSeverity none
redteam01ClaimInflationOutcome caught
redteam01ClaimInflationGrades VERIFIED 5, NOT_FOUND 8
redteam02ParaphraseDriftCases 12
redteam02ParaphraseDriftWorks 7
redteam02ParaphraseDriftRate 0.5833
redteam02ParaphraseDriftSeverity LOW
redteam02ParaphraseDriftOutcome works
redteam02ParaphraseDriftGrades VERIFIED 4, PARTIAL 1, UNVERIFIABLE 6
redteam03CounterpartyConfusablesCases 17
redteam03CounterpartyConfusablesWorks 1
redteam03CounterpartyConfusablesRate 0.0588
redteam03CounterpartyConfusablesSeverity MEDIUM
redteam03CounterpartyConfusablesOutcome works
redteam03CounterpartyConfusablesGrades VERIFIED 1, PARTIAL 4, NOT_FOUND 12
redteam04TimeTricksCases 7
redteam04TimeTricksWorks 0
redteam04TimeTricksRate 0
redteam04TimeTricksSeverity none
redteam04TimeTricksOutcome caught
redteam04TimeTricksGrades PARTIAL 1, NOT_FOUND 3
redteam05DomSpoofingCases 5
redteam05DomSpoofingWorks 0
redteam05DomSpoofingRate 0
redteam05DomSpoofingSeverity none
redteam05DomSpoofingOutcome caught
redteam05DomSpoofingGrades VERIFIED 1, NOT_FOUND 2, UNVERIFIABLE 2
redteam06PromptInjectionCases 4
redteam06PromptInjectionWorks 0
redteam06PromptInjectionRate 0
redteam06PromptInjectionSeverity none
redteam06PromptInjectionOutcome caught
redteam06PromptInjectionGrades VERIFIED 2, PARTIAL 2
redteam07CoverageLiesCases 7
redteam07CoverageLiesWorks 1
redteam07CoverageLiesRate 0.1429
redteam07CoverageLiesSeverity MEDIUM
redteam07CoverageLiesOutcome works
redteam07CoverageLiesGrades VERIFIED 1, NOT_FOUND 2, UNVERIFIABLE 4
redteam08ObserverStarvationCases 5
redteam08ObserverStarvationWorks 1
redteam08ObserverStarvationRate 0.2
redteam08ObserverStarvationSeverity LOW
redteam08ObserverStarvationOutcome works
redteam08ObserverStarvationGrades VERIFIED 2, NOT_FOUND 1, UNVERIFIABLE 2
redteam09DuplicateClaimsCases 4
redteam09DuplicateClaimsWorks 0
redteam09DuplicateClaimsRate 0
redteam09DuplicateClaimsSeverity none
redteam09DuplicateClaimsOutcome caught
redteam09DuplicateClaimsGrades VERIFIED 5
redteam10UnaskedEvasionCases 25
redteam10UnaskedEvasionWorks 18
redteam10UnaskedEvasionRate 0.72
redteam10UnaskedEvasionSeverity HIGH
redteam10UnaskedEvasionOutcome works
redteam10UnaskedEvasionGrades NOT_FOUND 3, UNASKED 7, UNVERIFIABLE 32
redteam11GrantBypassCases 10
redteam11GrantBypassWorks 0
redteam11GrantBypassRate 0
redteam11GrantBypassSeverity none
redteam11GrantBypassOutcome caught
redteam11GrantBypassGrades OVER_SCOPE 9, UNVERIFIABLE 7
redteam14ForgedEvidenceCases 5
redteam14ForgedEvidenceWorks 0
redteam14ForgedEvidenceRate 0
redteam14ForgedEvidenceSeverity none
redteam14ForgedEvidenceOutcome caught
redteam14ForgedEvidenceGrades PARTIAL 1, UNVERIFIABLE 4
redteam15UnaskedFalseAccusationCases 25
redteam15UnaskedFalseAccusationWorks 6
redteam15UnaskedFalseAccusationRate 0.24
redteam15UnaskedFalseAccusationSeverity MEDIUM
redteam15UnaskedFalseAccusationOutcome works
redteam15UnaskedFalseAccusationGrades UNASKED 6, UNVERIFIABLE 37
commit 23ea9c6b689b37f3ab74dd64ea2c70d16177e754
commitDirty false
node v22.14.0

ui

Measured 2026-10-13 with pnpm bench.

Metric Value
uiReceipts 5000
uiRuns 5
uiPopupItems 50
uiPopupReceiptRowsRead 204
uiPopupFullScans 0
uiPopupMedianMs 16.2
uiPopupMinMs 11.6
uiPopupMaxMs 17.2
uiPopupCpuMedianMs 17.7
uiTimelineCpuMedianMs 13.4
uiFilteredCpuMedianMs 23.2
uiDetailCpuMedianMs 0.8
uiGateMs 50
uiGateMetric uiPopupCpuMedianMs
uiGatePassed true
uiTimelineGatePassed true
commit 23ea9c6b689b37f3ab74dd64ea2c70d16177e754
commitDirty false
node v22.14.0