research-document FE-BND-FE-CONSOLIDATION-001

FE candidate consolidation log

FE candidate consolidation log

The ten starting candidates remain represented by eight A-series cards. The frozen v1.0 consolidation is retained because independent reconstruction found no evidence that the operations or predicted effects require a different grouping.

Starting candidate v1.1 card Curator decision Reason
Identity-capability separation A01 Retain Has its own classification operation and reviewer-agreement prediction.
Structured extraction and redesign A02 Retain Has a redesign operation, masked task-output test, and explicit complexity tradeoff.
Finite procedural-reasoning primitives A03 Merge with compositional grammar The vocabulary is tested through normalization, composition, reconstruction, and transfer; current records do not supply a separate outcome for mere finiteness.
Compositional procedural-reasoning grammar A03 Merge with primitive vocabulary Composition and reconstruction are the operational test of the vocabulary claim.
Reasoning grammar separated from execution or coordination grammar A04 Retain Has a separate artifact-family prediction and a weakening test, although the mechanism remains proposal-heavy.
Preserve observation, interpretation, uncertainty, and alternatives A05 Retain Has governance rules and a traceability prediction distinct from reassessment timing.
Verification and reassessment triggers A06 Retain Triggered reopening and verification are temporally different from statement typing, but the prediction and metrics are absent.
Framework-specific provenance, frozen instruments, and certified inputs A07 Merge with identifiers and history In the current research pipeline, IDs, snapshots, digests, normalization certificates, and frozen instruments jointly implement run lineage.
Canonical identifiers, supersession, and append-oriented history A07 Merge with provenance controls Current records operationalize identifiers and frozen-run lineage; broad supersession/history behavior remains incomplete rather than forming a second mechanism.
Characterization through test workflow A08 Retain The multi-stage governance flow exists, but no bundle interaction prediction is specified.

Decisions and uncertainty

  • No new candidate was added.
  • No candidate was removed because of weak evidence.
  • A04, A06, and A08 remain explicitly weak rather than being repaired with curator-created causal claims.
  • A07 records the absence of a comprehensive supersession policy.
  • No adjacent-method material or mapping hypothesis was consulted.