vocabulary-rule

To normalize ambiguous vocabulary across governed documents, apply a registry-driven baseline that maps terms to canonical meanings, decision rules, and allowed values. This removes ambiguity from global_instructions, AGENTS, and SKILL.md-family files before agents take action.

Resolving conflicting rule-strength expressions like must and should requires establishing strength semantics that prescribe priority resolution across global_instructions, AGENTS, and SKILL.md-family documents, ensuring agents interpret behavioral expectations consistently.

Maintain an active vocabulary registry by defining canonical term meanings, decision rules, allowed values, and prohibited interpretations. This registry normalizes ambiguous terms and provides a single enforceable baseline for authoring and reviewing governed documents.

Yes, you can track history and audit changes to term meanings by defining history recording paths and review workflows. This ensures that changes to vocabulary interpretations and rule semantics are tracked and verifiable during governance processes.

Vocabulary normalization works directly with global_instructions, CLAUDE, AGENTS, and SKILL.md-family documents. It applies to authoring, review, runtime interpretation, and reporting workflows to enforce consistent term meanings and rule-strength expressions.

Use vocabulary normalization when an agent encounters ambiguous instructions like "should prioritize soon" and needs to convert the phrase into a defined priority, record rationale for deviations, and log the change in the configured history path.