Fix the footer leaking "wrapper. #}" from a broken Jinja comment (#211)

The note explaining the hidden digest form contained a literal "{# … #}" in its
prose. Jinja doesn't nest comments, so the inner "#}" closed the comment early and
the trailing "wrapper. #}" rendered as visible text at the top of every footer.

Reworded the note without comment delimiters. Adds a test asserting no base-
template page's footer contains a stray "{#"/"#}" (fails against the old template).

Claude-Session: https://claude.ai/code/session_013dRZmX9D3JEntfMKWMTWZ8

templates/base.html.j2

@@ -72,8 +72,8 @@
 
   <footer class="site-footer">
     {# The compact digest signup is the site-wide footer pattern, not a
-       homepage-only surface — every page is a possible entry point.
-       Hidden for now; restore by removing the surrounding {# … #} wrapper. #}
+       homepage-only surface: every page is a possible entry point.
+       Hidden for now; restore by uncommenting the form block below. #}
     {#
     <form class="digest-form digest-compact" method="post" action="{{ base }}/digest"
           data-digest aria-labelledby="footer-digest-label">

tests/test_content.py

@@ -197,6 +197,18 @@ def test_hub_glossary_about(client):
     assert client.get(f"{B}/about").status_code == 200
 
 
+def test_footer_has_no_leaked_jinja_comment(client):
+    # A comment whose prose contained a literal "#}" closed early and leaked
+    # "wrapper. #}" into the footer. Guard every base-template page against a
+    # stray Jinja comment delimiter reaching the rendered HTML.
+    import re
+    for path in (f"{B}/about", f"{B}/climate/{SLUG}", f"{B}/climate/{SLUG}/records"):
+        b = client.get(path).text
+        footer = re.search(r"<footer.*?</footer>", b, re.S)
+        assert footer, f"no footer on {path}"
+        assert "#}" not in footer.group(0) and "{#" not in footer.group(0), path
+
+
 def test_title_name_is_short_and_unique():
     # Titles lead with the bare city name so the payload ("… in July") survives
     # SERP truncation — display_name()'s region+country is what pushed it off the