diff --git a/frontend/AGENTS.md b/frontend/AGENTS.md
index 893332f769..57e0c7c722 100644
--- a/frontend/AGENTS.md
+++ b/frontend/AGENTS.md
@@ -117,8 +117,9 @@ Read [src/AGENTS.md](src/AGENTS.md) for component and domain ownership.
 ## Code Style
 
 `core/utils/markdown.ts` reads web-fetch titles from the first nonblank line.
-Match zero to three literal spaces before `# ` without trimming indentation;
-mixed space/tab code blocks must fall back to the URL. Strip optional closing
+Match zero to three literal spaces before a single `#` followed by an ASCII
+space or tab, without trimming indentation; mixed space/tab code blocks must
+fall back to the URL. Strip optional closing
 hashes with a backwards scan so long fetched titles cannot cause quadratic regex
 backtracking. Closing hashes require a preceding space/tab (or an empty title)
 and only ASCII spaces/tabs afterward, apart from the terminal CR in CRLF input.
diff --git a/frontend/src/core/utils/markdown.ts b/frontend/src/core/utils/markdown.ts
index 24697c1ab2..59383ce4a4 100644
--- a/frontend/src/core/utils/markdown.ts
+++ b/frontend/src/core/utils/markdown.ts
@@ -5,7 +5,7 @@ export function extractTitleFromMarkdown(markdown: string) {
   if (firstLine === undefined) {
     return undefined;
   }
-  const headingPrefix = /^ {0,3}# /.exec(firstLine);
+  const headingPrefix = /^ {0,3}#[ \t]/.exec(firstLine);
   if (!headingPrefix) {
     return undefined;
   }
diff --git a/frontend/tests/unit/core/utils/markdown.test.ts b/frontend/tests/unit/core/utils/markdown.test.ts
index ce77694833..91f4737edc 100644
--- a/frontend/tests/unit/core/utils/markdown.test.ts
+++ b/frontend/tests/unit/core/utils/markdown.test.ts
@@ -6,6 +6,27 @@ test("reads the title from a leading ATX heading", () => {
   expect(extractTitleFromMarkdown("# Real Title\n\nbody")).toBe("Real Title");
 });
 
+test.each([
+  ["#\tReal Title", "Real Title"],
+  ["#\t Real Title", "Real Title"],
+  ["# \tReal Title", "Real Title"],
+  ["   #\tReal Title\r\nbody", "Real Title"],
+  ["#\tReal Title ###\t\r\nbody", "Real Title"],
+  ["#\t", undefined],
+  ["#\t###", undefined],
+])("accepts ASCII tab separators in %j", (markdown, expected) => {
+  expect(extractTitleFromMarkdown(markdown)).toBe(expected);
+});
+
+test.each(["\u00a0", "\u2003", "\v", "\f"])(
+  "does not treat %j as an ATX separator",
+  (separator) => {
+    expect(
+      extractTitleFromMarkdown(`#${separator}Not A Title`),
+    ).toBeUndefined();
+  },
+);
+
 test.each([
   ["# Release Notes ###", "Release Notes"],
   ["# Release Notes ###   ", "Release Notes"],