Examples should make decisions visible. A useful case study shows the original problem, constraints, route and content architecture, design direction, agent workflow, deployment target, and the checks that actually ran.
This section distinguishes three evidence levels: real customer work, internal experiments, and fictional demonstrations. Brief & Ship does not currently publish a customer case study here. It will not turn an invented brief into an invented testimonial.
Current documented build
The first case study is this marketing site. It documents the content collections, local assets, design tokens, metadata, Cloudflare configuration, and checks that can be inspected in the repository. Controlled experiments belong in Labs only after their prompts, environments, outputs, and scoring material exist.
Requirements for the next example
A future fictional demonstration must be labeled before the first paragraph and avoid customer language. An internal experiment must publish its method. A real customer case requires permission and support for every quoted outcome.
Those rules reduce the number of pages we can publish quickly. That is the point. The production-gap guide explains why credible release evidence matters more than a polished screenshot.