Documentation article
Understand loss analysis
Use diagnostics to understand where semantics degrade or remain unsupported.
Loss analysis is how we stay honest about transformations between formats that do not share the same semantics.
A successful conversion is not the same thing as a perfect conversion. The docs should make that distinction obvious early.
Loss happens when the source format can express meaning that the target format cannot encode cleanly, or when the current generator does not yet model that meaning.
The toolkit should surface those cases as diagnostics instead of hiding them inside a best-effort output.
This is especially important for downstream teams that may otherwise assume the generated result is safe to commit without review.
As the docs mature, this topic should evolve into concrete matrices and examples per route. For now, the goal is to set the right expectation and make the language consistent across the app.
