Walmsles

I have noticed that the Claude model sometimes creates markdown files with linting issues, which prompted me to create these rules, extracted from the markdown-lint code on GitHub.

Some rules defining strict ways of working, encouraging specifications to be completed, and a detailed development plan including AI prompts. The main components are enforcing a TDD approach where tests are written first and any test failures means we need to focus on the code (not change the tests).

Encourages a Test-Driven Design approach, with behaviour-like descriptions. The output of the test run should resemble a specification. Tests will be co-located with the actual code files and follow the *.spec.ts naming convention. The final rule is Definition of Done, which prompts the AI assistant to ensure that the code is in a good state with all tests passing and no errors or warnings. Note: Linting examples assume npm lint:fix script exists.