law-reverse-validate

Validating machine_readable law sections involves checking that every element is grounded in the original article text to prevent hallucinations. This process enforces scope fidelity, verifies correct source references, and ensures alignment with article boundaries during corpus generation or YAML auditing.

Hallucinations in machine_readable regulation YAML occur when generated elements lack grounding in the original legal text. This happens during corpus generation or editing when inputs, definitions, or conditions drift beyond their intended article boundaries or violate source-based design dependencies.

Auditing machine_readable legal data for scope fidelity should occur after creating or editing machine_readable sections, during corpus generation, or when auditing YAML data for legal accuracy. This ensures every element traces correctly to the original article text.

Detecting cross-provision dependencies in machine_readable law requires checking that all inputs, definitions, and conditions strictly reference their parent article text. The validation process flags cross-law dependencies that violate source-based design principles.

Source-grounding validation can be applied to existing legal YAML files during auditing tasks. It checks that every machine_readable element traces to the original article text, enforcing correct use of source references and alignment with article boundaries.

Validation of machine_readable law against source text is limited to checking scope fidelity, source reference correctness, and article boundary alignment. It does not generate new content but verifies existing machine_readable sections remain grounded in the original legal text.