Domain Modeling covers changing a project’s domain language; reading GLOSSARY.md for vocabulary is a one-line habit any skill can do. The skill challenges terms that conflict with the glossary, proposes a canonical term for fuzzy ones, invents edge-case scenarios, checks the code against what you say, and writes each resolved term down immediately.
It also decides when an architecture decision record is worth writing.
What You Get
- A GLOSSARY.md in a fixed format: bold term, a one or two sentence definition and an Avoid line listing rejected synonyms
- A GLOSSARY-MAP.md for multi-context repos, listing the contexts and the relationships between them
- Sequentially numbered ADRs in docs/adr, each as short as one paragraph
- Contradictions between your statements and the code, raised as questions
Common Use Cases
- Settling whether account means Customer or User before naming modules
- Recording why a database or messaging pattern was chosen so that no one reverses it unknowingly
- Splitting one glossary into per-context glossaries as a system grows
What Makes It Unique
It sets a three-part test for ADRs: the decision is hard to reverse, surprising without context and the result of a real trade-off. If any part is missing, no ADR is offered. It also keeps the glossary free of implementation details and general programming concepts, and creates files only when there is something to write.
Under The Hood
The skill is Markdown with two format guides, GLOSSARY-FORMAT.md and ADR-FORMAT.md. It runs no scripts. During a session the model reads the code, and creates or edits GLOSSARY.md, GLOSSARY-MAP.md and docs/adr files in your repository; those file writes are its only side effect.