Fix pages rendering: capture rendered markdown before MergeContent replaces it

MergeContent swaps the document content with the Razor template, so
GetContentStringAsync() in the model builder returned Razor source instead
of the rendered markdown. Store the HTML in RenderedBody metadata after
RenderMarkdown, then read it from there when building the Page model.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

Author: petrsvihlik

Pipelines/PagesPipeline.cs

@@ -19,6 +19,7 @@ public PagesPipeline()
                 new ReadFiles("pages/*.md"),
                 new ExtractFrontMatter(new ParseYaml()),
                 new RenderMarkdown(),
+                new SetMetadata("RenderedBody", Config.FromDocument(async doc => await doc.GetContentStringAsync())),
                 new SetDestination(Config.FromDocument(doc => GetPath(doc))),
             };
 
@@ -33,7 +34,7 @@ public PagesPipeline()
                         {
                             Title = document.GetString("title"),
                             Url = slug,
-                            Body = document.GetContentStringAsync().GetAwaiter().GetResult(),
+                            Body = document.GetString("RenderedBody"),
                             MetaDescription = document.GetString("description"),
                             ShowInNavigation = document.GetBool("show_in_navigation"),
                         };