fix(markdown): preserve links in list items

src/note_mcp/utils/html_to_markdown.py

@@ -313,6 +313,9 @@ def _convert_list(html_content: str, ordered: bool = False, indent_level: int =
             text = _OL_PATTERN.sub("", text)
             text = text.strip()
 
+        # Preserve supported inline elements before removing residual HTML tags.
+        text = _convert_inline_elements(text)
+
         # Clean up any remaining HTML tags from text
         text = re.sub(r"<[^>]+>", "", text).strip()
 

tests/unit/test_html_to_markdown.py

@@ -3,6 +3,8 @@
 Tests for html_to_markdown function that converts note.com HTML format back to Markdown.
 """
 
+from bs4 import BeautifulSoup
+
 from note_mcp.utils.html_to_markdown import html_to_markdown
 from note_mcp.utils.markdown_to_html import markdown_to_html
 
@@ -102,6 +104,47 @@ def test_nested_list_conversion(self) -> None:
         for line in sub_items:
             assert line.startswith("  ") or line.startswith("    ")
 
+    def test_ordered_and_unordered_list_links_are_preserved(self) -> None:
+        """Test that links survive conversion from paragraph-wrapped list items."""
+        for tag, marker in (("ul", "-"), ("ol", "1.")):
+            html = f"""<{tag}>
+                <li><p>Read the <a href="https://github.com/drillan/note-mcp">official docs</a>.</p></li>
+            </{tag}>"""
+
+            result = html_to_markdown(html)
+
+            assert f"{marker} Read the [official docs](https://github.com/drillan/note-mcp)." in result
+
+    def test_nested_list_links_preserve_formatted_labels(self) -> None:
+        """Test links and inline formatting in nested list items."""
+        reference_url = "https://note.com/drillan/n/n7379c02632c9"
+        html = f"""<ul>
+            <li><p>Start with <a href="https://github.com/drillan/note-mcp"><strong>API docs</strong></a>.</p>
+                <ol>
+                    <li><p>Then read the <a href="{reference_url}"><em>reference guide</em></a>.</p></li>
+                </ol>
+            </li>
+        </ul>"""
+
+        result = html_to_markdown(html)
+
+        assert "- Start with [**API docs**](https://github.com/drillan/note-mcp)." in result
+        assert f"  1. Then read the [*reference guide*]({reference_url})." in result
+
+    def test_list_link_roundtrip_preserves_label_and_href(self) -> None:
+        """Test that a list link keeps its label and href through HTML and Markdown."""
+        source_html = """<ol>
+            <li><p>Consult <a href="https://github.com/drillan/note-mcp"><strong>the guide</strong></a>.</p></li>
+        </ol>"""
+
+        markdown = html_to_markdown(source_html)
+        roundtrip_html = markdown_to_html(markdown)
+        links = BeautifulSoup(roundtrip_html, "html.parser").select('li a[href="https://github.com/drillan/note-mcp"]')
+
+        assert "1. Consult [**the guide**](https://github.com/drillan/note-mcp)." in markdown
+        assert len(links) == 1
+        assert links[0].get_text() == "the guide"
+
     # === Blockquote Tests ===
 
     def test_blockquote_conversion(self) -> None: