← ClaudeAtlas

ontologieslisted

Use when building or updating a project's domain vocabulary. Triggered by 'ontologies', 'build vocabulary', 'name things', 'define terms', 'ONTOLOGIES.md', or any request to establish precise domain language for a codebase.
tomcounsell/ai · ★ 14 · AI & Automation · score 73
Install: claude install-skill tomcounsell/ai
# Skill: /ontologies ## Purpose Build and maintain an `ONTOLOGIES.md` at the repo root — a canonical domain vocabulary that prevents naming confusion, documents bounded contexts, and surfaces when a single name is doing two jobs. ## When to Use - A new domain concept has appeared and the team is using different names for it - A code review surfaces confusion about what "X" means vs "Y" - A plan doc uses terms that are undefined or inconsistently applied - Before naming a new module, class, or API field — check existing terms first - When the user says "ontologies", "define this term", "what do we call X", or "ONTOLOGIES.md" ## Steps 1. **Read existing artifacts first.** Before asking any questions: - Read `ONTOLOGIES.md` at the repo root (if it exists) - Read `docs/adr/` for architectural decisions that defined terms - Grep for the term being discussed: `grep -r "<term>" --include="*.py" --include="*.md" -l` 2. **Identify the term(s) to clarify.** If invoked with no argument, ask: "What term or concept should we add to the ontology?" Wait for the answer. 3. **Run the interview loop via /grill-me.** Tell the user: "Running /grill-me on this term to surface the definition precisely." Ask one question at a time: - "What does this concept represent in the domain — not in code?" - "What is an example of this in production?" - "What is it NOT — what would a confused person mistake it for?" - "Is there a related concept that should be contrasted?" - "Do