From 4d01839bd3850988b8ecfaad13b198e5b665b8de Mon Sep 17 00:00:00 2001
From: konard <drakonard@gmail.com>
Date: Tue, 14 Jul 2026 08:39:13 +0000
Subject: [PATCH 1/3] Initial commit with task details

Adding .gitkeep for PR creation (default mode).
This file will be removed when the task is complete.

Issue: https://github.com/xlabtg/teleton-agent/issues/711
---
 .gitkeep | 1 +
 1 file changed, 1 insertion(+)
 create mode 100644 .gitkeep

diff --git a/.gitkeep b/.gitkeep
new file mode 100644
index 00000000..793e1c4c
--- /dev/null
+++ b/.gitkeep
@@ -0,0 +1 @@
+# .gitkeep file auto-generated at 2026-07-14T08:39:13.382Z for PR creation at branch issue-711-fb6ee98f2eb1 for issue https://github.com/xlabtg/teleton-agent/issues/711
\ No newline at end of file

From baeb4e4742bb94e9b5111fb7a836e74236b9c5ce Mon Sep 17 00:00:00 2001
From: konard <drakonard@gmail.com>
Date: Tue, 14 Jul 2026 08:56:47 +0000
Subject: [PATCH 2/3] =?UTF-8?q?test(sanitize):=20=D0=B2=D0=BE=D1=81=D0=BF?=
 =?UTF-8?q?=D1=80=D0=BE=D0=B8=D0=B7=D0=B2=D0=B5=D1=81=D1=82=D0=B8=20=D0=BF?=
 =?UTF-8?q?=D0=BE=D1=82=D0=B5=D1=80=D1=8E=20=D1=82=D0=B5=D0=BA=D1=81=D1=82?=
 =?UTF-8?q?=D0=B0=20=D0=BF=D0=BE=D1=81=D0=BB=D0=B5=20=D1=82=D0=B5=D0=B3?=
 =?UTF-8?q?=D0=B0?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

---
 src/utils/__tests__/sanitize.test.ts | 9 +++++++++
 1 file changed, 9 insertions(+)

diff --git a/src/utils/__tests__/sanitize.test.ts b/src/utils/__tests__/sanitize.test.ts
index 1936d928..8822b1fa 100644
--- a/src/utils/__tests__/sanitize.test.ts
+++ b/src/utils/__tests__/sanitize.test.ts
@@ -265,6 +265,11 @@ describe("sanitizeForPrompt", () => {
       expect(sanitizeForPrompt(input)).toBe("3 < 5 and 7 > 2");
     });
 
+    it("should preserve plain text after a tag when it is followed by >", () => {
+      expect(sanitizeForPrompt("<a>foo>bar")).toBe("foo>bar");
+      expect(sanitizeForPrompt("<b>hello</b> world")).toBe("hello world");
+    });
+
     it("should handle malformed tags", () => {
       const input = "Text <incomplete";
       expect(sanitizeForPrompt(input)).toBe("Text <incomplete");
@@ -821,6 +826,10 @@ describe("sanitizeForContext", () => {
 });
 
 describe("sanitizeTaskDescription", () => {
+  it("should preserve plain text after a tag when it is followed by >", () => {
+    expect(sanitizeTaskDescription("<a>foo>bar")).toBe("foo>bar");
+  });
+
   it("should remove fragmented nested script tags", () => {
     const input = "<scrip<script>t>alert(1)</scrip<script>t>";
     const result = sanitizeTaskDescription(input);

From d79bedc0ee5e29c997154558ef2e35f5ad034d26 Mon Sep 17 00:00:00 2001
From: konard <drakonard@gmail.com>
Date: Tue, 14 Jul 2026 08:59:48 +0000
Subject: [PATCH 3/3] =?UTF-8?q?fix(sanitize):=20=D1=81=D0=BE=D1=85=D1=80?=
 =?UTF-8?q?=D0=B0=D0=BD=D1=8F=D1=82=D1=8C=20=D1=82=D0=B5=D0=BA=D1=81=D1=82?=
 =?UTF-8?q?=20=D0=BF=D0=BE=D1=81=D0=BB=D0=B5=20=D0=B7=D0=B0=D0=BA=D1=80?=
 =?UTF-8?q?=D1=8B=D1=82=D0=B8=D1=8F=20=D1=82=D0=B5=D0=B3=D0=B0?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

---
 .gitkeep              |  1 -
 src/utils/sanitize.ts | 16 +++++++++-------
 2 files changed, 9 insertions(+), 8 deletions(-)
 delete mode 100644 .gitkeep

diff --git a/.gitkeep b/.gitkeep
deleted file mode 100644
index 793e1c4c..00000000
--- a/.gitkeep
+++ /dev/null
@@ -1 +0,0 @@
-# .gitkeep file auto-generated at 2026-07-14T08:39:13.382Z for PR creation at branch issue-711-fb6ee98f2eb1 for issue https://github.com/xlabtg/teleton-agent/issues/711
\ No newline at end of file
diff --git a/src/utils/sanitize.ts b/src/utils/sanitize.ts
index 100e8645..5b7d8931 100644
--- a/src/utils/sanitize.ts
+++ b/src/utils/sanitize.ts
@@ -13,13 +13,15 @@ function stripMarkupTags(text: string): string {
     if (text[i] === "<" && isTagStart(text, i)) {
       let end = text.indexOf(">", i + 1);
       if (end !== -1) {
-        let next = end + 1;
-        while (next < text.length) {
-          const fragmentEnd = text.indexOf(">", next);
-          if (fragmentEnd === -1) break;
-          if (!isTagNameFragment(text.slice(next, fragmentEnd))) break;
-          end = fragmentEnd;
-          next = end + 1;
+        if (text.slice(i + 1, end).includes("<")) {
+          let next = end + 1;
+          while (next < text.length) {
+            const fragmentEnd = text.indexOf(">", next);
+            if (fragmentEnd === -1) break;
+            if (!isTagNameFragment(text.slice(next, fragmentEnd))) break;
+            end = fragmentEnd;
+            next = end + 1;
+          }
         }
         i = end;
         continue;