A01 — Identity-capability separation

• Input [direct]: A knowledge artifact and evidence about its dominant purpose and functions.
• Operation [direct]: Assign identity from dominant purpose; record functions separately as capabilities; do not conflate the two. Record composition only when reasoning structure exists.
• Output [direct]: Separate identity and capability descriptions; composition and domain may also be recorded.
• Scope conditions [direct]: Knowledge-artifact characterization, especially where capabilities overlap across artifact identities.
• Failure conditions [direct]: Capabilities become indistinguishable from identity, or reviewer disagreement does not improve.
• Predicted effect [direct]: Separation should improve classification consistency and reduce reviewer disagreement.
• Measurement [direct]: Reviewer agreement and ambiguity under the separated instrument.
• Named baseline [inference]: A single-label or undivided characterization instrument.
• Boundary condition [inference]: Dominant purpose and functions must be recoverable from the available artifact evidence.
• Implementation test [inference]: Independent reviewers characterize the same artifacts with separated and undivided instruments; the candidate fails if agreement and ambiguity do not improve.
• Decision rights [absent]: Canonical records do not assign who resolves disputed identity.
• Uncertainty handling [inference]: Preserve unresolved identity or capability assignments rather than forcing a label.
• Verification [direct]: Larger blind reviewer comparison is required.
• Provenance [inference]: Retain artifact evidence supporting each identity and capability entry.
• Implementation cost [absent]: Not estimated in current records.
• Limitations [direct]: Pilot-derived hypothesis; larger independent review and comparison with alternative schemes remain required.
• Fidelity check [inference]: Identity is based on dominant purpose, capabilities remain separately listed, and composition is not invented.
• Null-capable future experiment [direct/inference]: The prescribed independent comparison can return no agreement improvement or no ambiguity reduction.

Statement basis

• Direct: Rules, hypothesis, predicted agreement effect, success/failure criteria, and need for a larger blind comparison.
• Inference: Baseline wording, boundary, implementation procedure, uncertainty treatment, fidelity check, and item-level provenance.
• Proposal-only: None.
• Absent: Dispute authority and implementation-cost estimate.