diff --git a/src/mail/screens/LinkPanel.py b/src/mail/screens/LinkPanel.py
index 08e2160..4ef598a 100644
--- a/src/mail/screens/LinkPanel.py
+++ b/src/mail/screens/LinkPanel.py
@@ -86,6 +86,9 @@ class LinkItem:
         - Keeping first and last path segments, eliding middle only if needed
         - Adapting to available width
         """
+        # Nerdfont chevron separator (nf-cod-chevron_right)
+        sep = " \ueab6 "
+
         # Special handling for common sites
         path = path.strip("/")
 
@@ -95,26 +98,26 @@ class LinkItem:
             if match:
                 repo, type_, num = match.groups()
                 icon = "#" if type_ == "issues" else "PR#"
-                return f"{domain} > {repo} {icon}{num}"
+                return f"{domain}{sep}{repo} {icon}{num}"
 
             match = re.match(r"([^/]+/[^/]+)", path)
             if match:
-                return f"{domain} > {match.group(1)}"
+                return f"{domain}{sep}{match.group(1)}"
 
         # Google Docs
         if "docs.google.com" in domain:
             if "/document/" in path:
-                return f"{domain} > Document"
+                return f"{domain}{sep}Document"
             if "/spreadsheets/" in path:
-                return f"{domain} > Spreadsheet"
+                return f"{domain}{sep}Spreadsheet"
             if "/presentation/" in path:
-                return f"{domain} > Slides"
+                return f"{domain}{sep}Slides"
 
         # Jira/Atlassian
         if "atlassian.net" in domain or "jira" in domain.lower():
             match = re.search(r"([A-Z]+-\d+)", path)
             if match:
-                return f"{domain} > {match.group(1)}"
+                return f"{domain}{sep}{match.group(1)}"
 
         # GitLab
         if "gitlab" in domain.lower():
@@ -122,7 +125,7 @@ class LinkItem:
             if match:
                 repo, type_, num = match.groups()
                 icon = "#" if type_ == "issues" else "MR!"
-                return f"{domain} > {repo} {icon}{num}"
+                return f"{domain}{sep}{repo} {icon}{num}"
 
         # Generic shortening - keep URL readable
         if len(url) <= max_len:
@@ -136,31 +139,31 @@ class LinkItem:
 
         # Try to fit the full path first
         full_path = "/".join(path_parts)
-        result = f"{domain} > {full_path}"
+        result = f"{domain}{sep}{full_path}"
         if len(result) <= max_len:
             return result
 
         # Keep first segment + last two segments if possible
         if len(path_parts) >= 3:
             short_path = f"{path_parts[0]}/.../{path_parts[-2]}/{path_parts[-1]}"
-            result = f"{domain} > {short_path}"
+            result = f"{domain}{sep}{short_path}"
             if len(result) <= max_len:
                 return result
 
         # Keep first + last segment
         if len(path_parts) >= 2:
             short_path = f"{path_parts[0]}/.../{path_parts[-1]}"
-            result = f"{domain} > {short_path}"
+            result = f"{domain}{sep}{short_path}"
             if len(result) <= max_len:
                 return result
 
         # Just last segment
-        result = f"{domain} > .../{path_parts[-1]}"
+        result = f"{domain}{sep}.../{path_parts[-1]}"
         if len(result) <= max_len:
             return result
 
         # Truncate with ellipsis as last resort
-        result = f"{domain} > {path_parts[-1]}"
+        result = f"{domain}{sep}{path_parts[-1]}"
         if len(result) > max_len:
             result = result[: max_len - 3] + "..."
 
diff --git a/src/mail/widgets/ContentContainer.py b/src/mail/widgets/ContentContainer.py
index 7552f4a..bfe6dfd 100644
--- a/src/mail/widgets/ContentContainer.py
+++ b/src/mail/widgets/ContentContainer.py
@@ -56,7 +56,7 @@ def compress_urls_in_content(content: str, max_url_len: int = 50) -> str:
 
         # Keep original anchor text, but if it's the same as URL, use short version
         if anchor_text == url or anchor_text.startswith("http"):
-            return f"[🔗 {short_url}]({url})"
+            return f"[\uf0c1 {short_url}]({url})"
         else:
             return match.group(0)  # Keep original if anchor text is meaningful
 
@@ -74,7 +74,7 @@ def compress_urls_in_content(content: str, max_url_len: int = 50) -> str:
         )
 
         # Return as markdown link with icon
-        return f"[🔗 {short_url}]({url})"
+        return f"[\uf0c1 {short_url}]({url})"
 
     # First, process markdown links
     md_link_pattern = r"\[([^\]]+)\]\((https?://[^)]+)\)"