Rule Model Foundations
Fact Input Record
Rule models need explicit fact inputs rather than hidden assumptions. Honesty note: simplified toy rule-modeling workflow, corpus-light with no neochart dependency; the pinned first step carries the full honesty note.
Structured Visual
Jurisdiction: US; as of 2026-06-24; not legal advice; Code encodes a stated interpretation workflow, not the law itself.
RENDER STRUCTURE · REFUSE INTERPRETATION · CITE · ABSTAIN · HAND-OFF: render structure, refuse interpretation, cite provenance, abstain when unsupported, and hand off to human review.
Rule model honesty note
Honesty note: simplified toy rule-modeling workflow; corpus-light and no neochart dependency; jurisdictions and interpretations vary; as of June 24, 2026; not legal advice; code encodes a stated interpretation workflow, not the law itself.
Use explicit fact fields
The stated process turns prose phrases into named fact-input fields with known or unknown status.
Example fact packet
The toy packet includes age years, residency scope, filing date, and exclusion flag fields.
The table recomputes support
The retrieval table recomputes 4 fields and 4 token-supported rows from the authored fact-input text.
Diagram note
The diagram is a fact-field table. It exposes inputs and unknowns instead of hiding assumptions.
Jurisdiction: US; as of 2026-06-24; not legal advice; Code encodes a stated interpretation workflow, not the law itself.
RENDER STRUCTURE · REFUSE INTERPRETATION · CITE · ABSTAIN · HAND-OFF: render structure, refuse interpretation, cite provenance, abstain when unsupported, and hand off to human review.
Summary
Fact records make the model reviewable because each input has a name, source, and status.