fix(agent): preserve preview URLs for collection repo models with flat heading structure

Three-part fix for enrich_hf_metadata failing to extract correct preview_url
from HuggingFace collection repos where models share flat heading levels:

1. _strip_standalone_images() now converts <img> tags to markdown image
   syntax ![alt](src) instead of stripping the URL entirely, so the LLM
   can still extract preview URLs.

2. _extract_section() uses a line-count-based forward window (stopping at
   <a id> anchors) for non-heading matches, instead of stopping at the
   very next heading. This prevents same-level sub-headings (# Download,
   # Trigger, # Sample prompt within a single model section) from
   truncating the window before sample images are included.

3. Post-processor preview fallback now filters gallery images to the
   model-specific README section before falling back to the repo-wide
   first image.

tests/agent_cli/test_agent_cli.py

@@ -803,10 +803,12 @@ pixel art sprite, game asset, transparent background
         assert "## Gallery" in result
         assert "Some text." in result
 
-    def test_html_img_tag_stripped(self):
+    def test_html_img_tag_converted_to_markdown_image(self):
+        """``<img>`` converted to ``![](src)``, preserving URL for LLM."""
         md = '## Preview\n<img src="https://cdn.hf.co/img.webp"></img>\n\nDescription.'
         result = self._clean(md)
-        assert "cdn.hf.co" not in result
+        assert "![](https://cdn.hf.co/img.webp)" in result
+        assert "cdn.hf.co" in result  # URL preserved for LLM extraction
         assert "Description." in result
 
     def test_inline_image_within_paragraph_preserved(self):

tests/agent_cli/test_readme_processor.py

@@ -242,11 +242,12 @@ After"""
         cleaned = R.clean_readme_for_llm(text)
         assert "./preview.png" in cleaned
 
-    def test_strips_html_img_tag(self, R):
-        """``<img src="...">`` → stripped."""
+    def test_converts_html_img_tag_to_markdown_image(self, R):
+        """``<img src="...">`` → ``![](src)`` preserving URL for LLM."""
         text = 'before\n<img src="logo.png">\nafter'
         cleaned = R.clean_readme_for_llm(text)
-        assert "logo.png" not in cleaned
+        assert "![](logo.png)" in cleaned
+        assert "logo.png" in cleaned  # URL preserved for LLM extraction
 
     def test_widget_stripped_frontmatter_preserved(self, R):
         """Widget YAML stripped but ``base_model:`` kept."""