Files
whetstone_DSL/specialists/data/generated/automatability_binary_train.tsv
Bill 15125dce25 Specialist fleet: pilot sweep results + automatability decomposition
Pilot sweep (all 4 gates, small_plus + medium):
  verification_type  84.3%  (medium, no_gain_stop_scaling)
  worker_type        80.8%  (small_plus, scale_more)
  automatability     78.3%  (medium, decompose_or_relabel)
  prereq_op          72.2%  (small_plus, decompose_or_relabel)

Automatability: 2-stage decomposition resolves Directional Collapse.
  Stage 1 binary (low_entropy vs needs_routing): 100%
  Stage 2 4-way (specialist/slm/llm/human): 100%
  Combined 5-way pipeline: 100%
  Script: specialists/scripts/split_automatability_decompose.py

Worker_type: combined dataset (sprint plans + projects) improves
  implementer class from 42.9% → 75.0%, overall 80.8% → 92.2%.
  Combined train/eval TSVs added.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
2026-04-19 13:44:38 -06:00

8.2 KiB

110name=AgentChatContextInjector
210name=ApprovalEscalationPlanner
310name=StrategyAwareOptimizer
400name=SettingsManager
510name=LibrarySymbolAdvisor
610name=TransformEngineExtended
710name=StrategyAwareOptimizer
810name=SymbolSelectorAPI
910name=ApprovalEscalationPlanner
1000name=PolyApiProject
1110name=MemoryModelTranslator
1210name=AgentChatSessionPersistence
1310name=AgentSessionRecorder
1410name=IncrementalOptimizer
1510name=TypeSystemTranslator
1600name=MigrationTestGenerator
1700name=SCIPEmitter
1810name=ArchitectReviewInterface
1900name=CapabilityDeclarationGenerator
2000name=Sprint190IntegrationSummary
2110name=SecurityExceptionReviewBoard
2200name=SecurityTestSkeletonGenerator
2300name=SchemaToCppGenerator
2400name=JobDispatchTableGenerator
2500name=Sprint85IntegrationSummary
2610name=SymbolSelectorAPI
2700name=ProjectAstCMakeGenerator
2800name=BatchMutationAPI
2910name=CanaryPromotionJudge
3010name=ArchitectReviewSurface
3110name=TypeAwareMappings
3210name=SecurityPreservingTranslation
3310name=TransformEngineExtended
3410name=ArchitectReviewSurface
3500name=Sprint222IntegrationSummary
3610name=ArchitectReviewInterface
3710name=ArchitectReviewSurface
3810name=PolyglotSuiteOrchestrator
3900name=Sprint265IntegrationSummary
4000name=SyntaxHighlighterOrg
4110name=TransformEngineExtended
4210name=TypeAwareMappings
4310name=AgentChatPanelModel
4410name=ArchitectModuleDecomposer
4500name=ContextSliceAssembler
4600name=SyntaxHighlighterJavaScript
4710name=TranspilationConfidence
4810name=ArchitectTechStackSelector
4910name=PerTaskLibrarySelector
5010name=ArchitectReviewInterface
5110name=SecurityExceptionReviewBoard
5200name=Sprint106IntegrationSummary
5310name=OperationClassifier
5400name=SQLiteDataLayerGenerator
5510name=SelfHostTranspileAudit
5600name=Sprint27IntegrationSummary
5710name=IncrementalOptimizer
5810name=AgentMutationApproval
5900name=Sprint98IntegrationSummary
6010name=OperationSelectorAPI
6110name=SelfHostTranspileAudit
6210name=OperationSelectorAPI
6300name=AccessibilityBaseline
6410name=AgentMutationApproval
6500name=TraceGenerator
6600name=CostEstimator
6710name=AgentPermissionPolicy
6810name=AgentMutationPreview
6910name=ConcurrencyTranslator
7010name=SelfHostTranspileAudit
7110name=ContextWindowOptimizer
7210name=AgentMutationPreview
7310name=IntentTranslator
7410name=AgentLibraryPolicy
7500name=SyntaxHighlighterPython
7600name=JobDispatchTableGenerator
7700name=EditorModePolicy
7810name=LibrarySymbolAdvisor
7910name=AgentMarketplace
8010name=TranspilationConfidence
8100name=SwarmStatusPanelModel
8210name=ApprovalEscalationPlanner
8310name=TransformEngine
8400name=SyntaxHighlighter
8510name=ArchitectIntakeProcessor
8610name=PolyglotOrchestrator
8710name=AgentTaskStatusOverlay
8810name=ArchitectReviewInterface
8910name=OperationSelectorAPI
9010name=LibrarySymbolAdvisor
9100name=Sprint184IntegrationSummary
9210name=SecurityExceptionReviewBoard
9300name=Sprint36aIntegration
9410name=ArchitectReviewInterface
9500name=SyntaxHighlighterPython
9600name=SearchUtils
9700name=Diagnostics
9800name=TestPlanGenerator
9910name=SecurityExceptionReviewBoard
10000name=InferenceJobGenerator
10110name=ArchitectProblemParser
10200name=LSPSymbolExtractor
10310name=SymbolSelectorAPI
10410name=ArchitectProjectPipeline
10510name=PerTaskLibrarySelector
10610name=AgentToolCallVisualization
10700name=ProjectSkeletonGenerator
10810name=TaskitemLibraryAnnotator
10910name=CanaryPromotionJudge
11010name=SecurityExceptionReviewBoard
11110name=OperationSelectorAPI
11200name=LanguageGeneratorPluginManifest
11310name=TransformEngine
11410name=OnCallCoveragePlanner
11510name=TaskitemLibraryAnnotator
11610name=TaskitemLibraryAnnotator
11710name=AgentMutationPreview
11810name=AgentChatSessionPersistence
11900name=CHeaderEmitter
12000name=Sprint49IntegrationSummary
12110name=ContextWindowOptimizer
12210name=PerTaskLibrarySelector
12310name=ArchitectMultiLanguageOrchestrator
12400name=MCPWorkflowPrompts
12510name=TranspilationRPC
12610name=ContextWindowOptimizer
12710name=OperationSelectorAPI
12800name=Sprint198IntegrationSummary
12910name=ArchitectScaffoldGenerator
13000name=GoCppBindingEmitter
13100name=AnimationUtils
13200name=SelfHostAnnotationAudit
13310name=ArchitectIntakeProcessor
13410name=AgentMarketplace
13510name=TranspilationConfidence
13610name=ArchitectReviewSurface
13700name=SecureByDefaultGenerator
13810name=OperationSelectorAPI
13910name=CanaryPromotionJudge
14010name=TaskitemLibraryAnnotator
14110name=AgentChatPanelModel
14210name=LibrarySymbolAdvisor
14310name=SecurityPreservingTranslation
14400name=APIBoundaryPreserver
14510name=ApprovalEscalationPlanner
14610name=TypeSystemTranslator
14710name=StrategyAwareOptimizer
14810name=TransformEngineExtended
14900name=SyntaxHighlighterRust
15010name=AgentSessionRecorder
15100name=Sprint136IntegrationSummary
15200name=OperationTaxonomy
15310name=ArchitectTechStackSelector
15400name=AnnotationConflict
15510name=CanaryPromotionJudge
15610name=ApprovalEscalationPlanner
15700name=Sprint286IntegrationSummary
15800name=MigrationPlanGenerator
15900name=Sprint261IntegrationSummary
16000name=TrainingDataGenerator
16100name=RustPythonBindingEmitter
16210name=AgentTaskSlots
16310name=AgentCodeGen
16400name=Sprint74IntegrationSummary
16500name=ArchitectTemplates
16610name=SymbolSelectorAPI
16710name=TaskitemLibraryAnnotator
16810name=ApprovalEscalationPlanner
16910name=CanaryPromotionJudge
17010name=AgentMutationApproval
17100name=SyntaxHighlighterJava
17210name=AgentMutationPreview
17310name=AgentMutationPreview
17400name=SQLiteDataLayerGenerator
17510name=AgentTaskSlots
17610name=AgentRPCHandler
17710name=MemoryModelTranslator
17810name=ConcurrencyTranslator
17910name=ArchitectReviewSurface
18010name=AgentToolCallVisualization
18110name=SecurityExceptionReviewBoard
18200name=RustGoBindingEmitter
18300name=Sprint58IntegrationSummary
18410name=IncrementalOptimizer
18500name=Sprint239IntegrationSummary
18600name=SyntaxHighlighterElisp
18700name=PlatformCMakeTargetGenerator
18810name=ApprovalEscalationPlanner
18900name=MqttBoilerplateGenerator
19010name=PerTaskLibrarySelector
19110name=IntentTranslator
19210name=SecurityExceptionReviewBoard
19300name=KeySymbolRenderer
19410name=OperationSelectorAPI
19510name=MemoryModelTranslator
19610name=ArchitectReviewInterface
19700name=ArchitectSkeletonGenerator
19810name=OperationClassifier
19910name=ApprovalEscalationPlanner
20010name=OperationClassifier
20100name=PackageScriptGenerator
20210name=TaskitemLibraryAnnotator
20310name=TypeAwareMappings
20410name=ArchitectReviewInterface
20500name=GoCppBindingEmitter
20610name=AgentMutationApproval
20700name=CHeaderEmitter
20800name=SyntaxHighlighterOrg
20900name=TaskitemGeneratorV2
21000name=Sprint96IntegrationSummary
21100name=ScoreClampUtil
21210name=AgentRPCHandler
21310name=ArchitectScaffoldGenerator
21410name=ScopeMilestoneDecomposer
21510name=AgentMutationApproval
21610name=AgentMutationPreview
21710name=AgentPermissionPolicy
21810name=PerTaskLibrarySelector
21900name=ConstrainedRoutingRulesetExtension
22000name=LanguageGeneratorPluginManifest
22110name=ArchitectReviewSurface
22200name=WorkItem
22310name=TaskitemLibraryAnnotator
22410name=CanaryPromotionJudge
22510name=AgentMutationApproval
22610name=LibrarySymbolAdvisor
22710name=ConcurrencyTranslator
22810name=CanaryPromotionJudge
22910name=WorkflowOrchestrator
23000name=InferenceJobGenerator
23110name=TranspilationRPC
23210name=AgentLibraryPolicy
23300name=Sprint272IntegrationSummary
23410name=TransformEngine
23510name=TranspilationRPC
23610name=OnCallCoveragePlanner
23710name=SymbolSelectorAPI
23810name=PolyglotOrchestrator
23910name=OperationClassifier
24000name=RefactorActions
24110name=LibrarySymbolAdvisor
24210name=IntentTranslator
24310name=PerTaskLibrarySelector
24410name=WorkflowOrchestrator
24510name=AgentChatContextInjector
24610name=AgentAnnotationAssistant
24710name=AgentMutationApproval
24810name=SymbolSelectorAPI
24910name=ArchitectReviewInterface
25000name=SyntaxHighlighterCpp
25110name=ArchitectMultiLanguageOrchestrator
25200name=APIDocGenerator
25310name=ArchitectProblemParser
25410name=StrategyAwareOptimizer
25510name=AgentAnnotationAssistant
25600name=PackageScriptGenerator
25710name=AgentRegistry
25810name=ArchitectModuleDecomposer
25900name=Sprint249IntegrationSummary
26010name=SecurityPreservingTranslation
26110name=ArchitectProjectPipeline
26200name=WorkerDispatchValidator
26310name=ContextWindowOptimizer
26410name=PolyglotSuiteOrchestrator
26510name=AgentMutationPreview
26610name=TypeSystemTranslator