What problem does it solve? Teams drift into using multiple words for the same domain concept across code, docs, and conversation, causing ambiguity and miscommunication. This Skill establishes one canonical vocabulary per project and keeps code, commits, and agents aligned with it. ## Core Features & Use Cases - Glossary scaffolding and interviews: Initialize docs/concepts/UBIQUITOUS_LANGUAGE.md, wire it into CLAUDE.md, then research the repo and interview the user to define canonical terms with Avoid synonym lists and Has relationships. - Validation and visualization: Run the bundled glossary.py script to lint entries (missing definitions, dangling relationships, canonical-vs-Avoid clashes) and generate a mermaid ER diagram from the relationships. - Drift auditing and renames: Scan code, commits, and docs for terms that conflict with the glossary, and perform safe canonical-term renames that retire old words to Avoid lists. - Use Case: A team notices 'Account' and 'Customer' used interchangeably in their codebase. Run the audit sub-command to surface the conflict, interview to pick the canonical term, then enforce it via a pre-commit hook. ## Quick Start Ask the agent to set up a ubiquitous language glossary for this project and interview you about the core domain terms.
