When Linting Shrugs: Dynamic Checks for API Specs

Static linting is an essential first line of defense for API specifications. It catches syntax errors, style violations, missing fields, and many structural issues before they escape into implementation. But semantic failures are also expensive: a schema may be valid and still fail to describe behavior that clients and servers can rely on.
This talk explores a more dynamic approach to validating API specs. Instead of only reading the spec as a document, we exercise it as an executable description of behavior using mocks, generated tests, boundary values, and examples. That forces subtle issues into the open: contradictory constraints, impractical limits, confusing content negotiation, ambiguous status codes, and missing error cases.
Through concrete examples and demos, we will compare what static linting catches with what dynamic validation reveals. We will also look at where this fits in an API delivery workflow so teams can use linting for fast structural feedback and dynamic validation for deeper confidence before specs reach consumers.