naming-register-policy - Two-register naming policy for machine codes and human-facing prose
Defines a two-register naming policy that classifies vocab-seed terms and rules for when machine codes stay as-is versus require human names or first-use expansion.
Tags
Updated: 2026-06-30Capabilities
Typical Inputs
Typical Outputs
What this skill does
- classify vocab-seed terms into register tiers
- resolve code token meanings in context
- decode numbered-family token instances
- define rename-candidate replacement worklist
- apply two-register naming rules to prose
- identify escape-hatch paths for codes
Inputs
- register.json asset file
- HOW-IT-WORKS.md §5 vocab-seed terms
- human-facing prose surfaces
- machine-citation context markers
Outputs
- term classification labels (stable-code, self-describing, rename-candidate)
- expansion text for stable-code terms
- replacement values for rename-candidate terms
- decode paths for family-grouped tokens
Requirements
- register.json asset file available in skills/naming-register-policy/assets/
- HOW-IT-WORKS.md §5 vocab-seed as canonical source
- Pester suite for bidirectional key-set validation