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>
887 B
887 B
| 1 | 0 | 0 | name=LibrarySymbolAdvisor |
|---|---|---|---|
| 2 | 1 | 0 | name=SelfHostTranspileAudit |
| 3 | 1 | 0 | name=IntentTranslator |
| 4 | 1 | 0 | name=TypeSystemTranslator |
| 5 | 1 | 0 | name=SecurityPreservingTranslation |
| 6 | 1 | 0 | name=IncrementalOptimizer |
| 7 | 3 | 0 | name=ApprovalEscalationPlanner |
| 8 | 3 | 0 | name=ArchitectReviewInterface |
| 9 | 1 | 0 | name=MemoryModelTranslator |
| 10 | 2 | 0 | name=AgentCodeGen |
| 11 | 0 | 0 | name=OperationClassifier |
| 12 | 0 | 0 | name=SymbolSelectorAPI |
| 13 | 3 | 0 | name=ArchitectReviewSurface |
| 14 | 0 | 0 | name=LibrarySymbolAdvisor |
| 15 | 0 | 0 | name=TaskitemLibraryAnnotator |
| 16 | 0 | 0 | name=PerTaskLibrarySelector |
| 17 | 2 | 0 | name=AgentRegistry |
| 18 | 0 | 0 | name=OperationClassifier |
| 19 | 3 | 0 | name=AgentMutationApproval |
| 20 | 3 | 0 | name=AgentMutationPreview |
| 21 | 1 | 0 | name=TransformEngine |
| 22 | 0 | 0 | name=OperationClassifier |
| 23 | 3 | 0 | name=ArchitectReviewSurface |
| 24 | 0 | 0 | name=OperationSelectorAPI |
| 25 | 3 | 0 | name=AgentMutationPreview |
| 26 | 3 | 0 | name=AgentMutationApproval |
| 27 | 2 | 0 | name=AgentTaskStatusOverlay |
| 28 | 2 | 0 | name=ScopeMilestoneDecomposer |
| 29 | 3 | 0 | name=SecurityExceptionReviewBoard |