This skill loads a compact rule set inspired by Robert C. Martin’s Clean Architecture. It tells a coding agent to point source dependencies inward, keep entities and use cases free of framework types, and treat databases, web delivery and vendors as replaceable details.
It applies to code whose business rules should survive changes in frameworks, storage, delivery mechanisms and deployment shape.
What You Get
- A mini rule set with decision rules, trigger rules and a closing checklist for architecture work
- A full version of the rules for deeper architectural tradeoffs
- A skill description that makes the agent load the rules for boundary, dependency and framework-isolation tasks
Common Use Cases
- Moving business branching out of controllers, jobs and handlers into use cases
- Introducing a policy-owned port in front of a database or service client
- Replacing a shared common package with structure organised by use case
- Reviewing a feature for framework types leaking into core logic
What Makes It Unique
The rules are written as trigger conditions, such as a use case that instantiates infrastructure or an adapter that bypasses a use case, each paired with the repair to make. They also ask for boundaries to be enforced by packages, build rules or tests, and for the lightest workable boundary when a full split costs too much. It sits beside the repository’s other architecture rule sets and focuses on dependency direction.
Under The Hood
The skill is a SKILL.md with a name and description, plus Markdown files holding the mini, nano and full versions of the rules. The SKILL.md tells the agent to read the mini file before design or code decisions and the full file only for harder tradeoffs. It contains no scripts. It does not fetch web pages, call APIs, run commands or write files; the agent only reads the rule text.