Snippet naming: content-derived on publish, frozen on explicit rename

docs/spec/02-snippet-library.md

@@ -87,6 +87,7 @@ The library provides the lifecycle operations for snippets. An operation whose o
 New snippets get a sensible default name, and a tag field exists on each snippet for categorization, though tags are not a primary user surface.
 
 - A new snippet receives an auto-generated default name based on the current date and time, so it is uniquely identifiable until the user renames it (renaming happens in the metadata panel).
+- Names follow a **provenance hierarchy**: an **explicitly chosen** name (set via rename in the metadata panel) is frozen — the app never rewrites it. Every **app-picked** name (the timestamp default, a Chart Builder-generated name, or a previously derived one) is a "next best pick" that keeps tracking the spec: on each publish it is re-derived as the spec's `title` when present, else a mark + encodings description (the same dialect the Chart Builder names its output in), else the existing name stands. So the library reads by chart rather than by creation time, until the user takes over a name — at which point their word is final (see _Spec Editor → Publish_, _Data Model → `nameSource`_).
 - Each snippet stores a list of **tags**. Tags are persisted and carried through duplication; for example, snippets brought in via import are tagged "imported" (see _Import & Export_).
 - There is no dedicated tag-management UI; tags are stored on the data model but are not surfaced as a primary browsing or editing control.